»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or rakudo:, std:, or /msg camelia p6: ... | irclog: irc.perl6.org | UTF-8 is our friend!
Set by masak on 12 May 2015.
00:11 ab6tract left 00:16 BenGoldberg joined 00:17 BenGoldberg left, BenGoldberg joined 00:26 lichtkind left, llfourn joined 00:27 xzenfar joined 00:30 llfourn left 00:50 yqt left 00:59 aborazmeh joined, aborazmeh left, aborazmeh joined 01:11 Sgeo left 01:12 Sgeo joined 01:13 xzenfar left 01:34 dha left 01:49 dayangkun joined 01:56 kaare__ joined
retupmoca YuviPanda: run: "panda gen-meta" from the root directory of your module to get a META.info created 01:59
YuviPanda: then you just need to add the link to that META.info to the META.list file in this repo: github.com/perl6/ecosystem
02:14 mdinger joined 02:18 noganex_ joined, yeahnoob joined 02:20 noganex left 02:27 llfourn joined 02:31 llfourn left 02:48 BenGoldberg left 03:09 [Sno] left 03:32 pierrot left 03:40 pierrot joined 03:41 xzenfar joined 03:45 mr-foobar joined 03:46 rangerprice left 03:49 mdinger left 04:22 llfourn joined 04:26 Calibellus left 04:31 skids left 04:32 telex left 04:34 telex joined 04:35 cognominal joined 05:08 davido_ joined 05:13 llfourn left 05:18 [Sno] joined
colomon needs to get moving so he can catch the 8:48 train to Olten. 05:21
05:29 breinbaas joined 05:30 davercc` joined 05:31 davercc left 05:32 khw left 05:33 dha joined, davido_ left 05:47 brrt joined 05:51 Sqirrel joined
colomon is it difficult to find the hackathon room from the front door of the venue? I’ve got the venue starred in google maps so I can find it, but I won’t have an easy way to communicate with IRC once I’m there.... 06:03
06:05 diana_olhovik_ joined
timotimo *yawn* 06:06
maettu timotimo: indeed.. :-)
06:07 llfourn joined
timotimo i forgot what people said when they were going to get breakfast, so i just guessed "getting up at 0800, then showering, then breaking fast shall be good enough" 06:08
maettu colomon: no, that's easy.
colomon maettu: yay!
maettu we're opening the door @ 9 am 06:09
colomon I should be there around 11
06:10 colomon left 06:11 bioexpress joined
masak morning, #perl6 06:18
hope everyone had a good night's sleep
masak heads down for breakfast
06:19 domidumont joined
bioexpress "doesnotexist" does not exist. Why dies 06:20
p6: my $contents = "doesnotexist".IO.slurp;
camelia rakudo-moar a7c845: OUTPUT«Failed to open file /home/camelia/doesnotexist: no such file or directory␤ in block <unit> at /tmp/tmpfile:1␤␤»
GLRelia rakudo-moar a11e8f: OUTPUT«Failed to open file /home/steve/doesnotexist: no such file or directory␤ in block <unit> at /tmp/tmpfile:1␤␤»
bioexpress and
p6: rmdir "dasnotexist";
GLRelia rakudo-moar a11e8f: OUTPUT«rmdir is disallowed in restricted setting␤ in sub restricted at src/RESTRICTED.setting:1␤ in sub rmdir at src/RESTRICTED.setting:13␤ in block <unit> at /tmp/tmpfile:1␤␤»
camelia rakudo-moar a7c845: OUTPUT«rmdir is disallowed in restricted setting␤ in sub restricted at src/RESTRICTED.setting:1␤ in sub rmdir at src/RESTRICTED.setting:13␤ in block <unit> at /tmp/tmpfile:1␤␤»
bioexpress does not die? 06:21
nine .tell skids Ah, yes thanks! I'll have a look at it after breakfast. Now that I understand the problem it solves, I think it's a shame, your PR has not been merged yet.
yoleaux nine: I'll pass your message to skids.
masak bioexpress: you seem surprised that `slurp` on something that doesn't exist dies. I'm not surprised at that. why are you? :) 06:22
bioexpress: also, removing files tends to not die if they're not there. kind of a vacuous success.
bioexpress: not entirely evident that this goes for directories as well as normal files, but...
06:23 domidumont left
bioexpress Thx! 06:24
06:26 abaugher left 06:27 abaugher joined 06:36 davercc` left
dalek kudo/glr: d44cd4b | skids++ | src/core/array_slice.pm:
Fix Whatevercode in array slice indices past the first one.

Lets us pass /S32-array/delete.t which used to fail with
  "Cannot call Real(WhateverCode: )"
  (Surprisingly this is the only file containing a test that uses them,
unless they are skipped)		 06:38
kudo/glr: 91d2127 | niner++ | src/core/array_slice.pm:
Merge pull request #515 from skids/glr3

Fix Whatevercode in array slice indices past the first one.
nine .tell skids thank you for fixing @a[1, *-1]:delete :) I've had a look at this yesterday but my train arrived before I came to writing code. I really wonder how this worked in nom though. Couldn't find any code handling that case there. 06:39
yoleaux nine: I'll pass your message to skids.
dha nine - have the 5to6 docs been of any use to you? 06:40
(it was you that was asking about them a while back, right?) 06:41
06:41 laben joined
laben hello p6'ers! 06:42
06:44 Sqirrel left
timotimo hello laben! 06:48
06:50 domidumont joined 06:52 Woodi left, Woodi joined 06:53 domidumont left, zakharyas joined 06:54 domidumont joined 06:56 brrt left
timotimo heads towards the breakfast area 07:04
moritz down to 84 failing test files, it seems
laben so is it under nine% ? 07:06
07:07 dha left 07:09 rurban joined 07:11 brrt joined 07:13 larion left 07:16 TEttinger left 07:20 darutoko joined 07:28 Sqirrel joined, laouji joined
TimToady waves from hackython 07:35
07:38 FROGGS left
nine Actually it's 83 here 07:39
07:42 espadrine joined
moritz might be one of the flappers 07:45
timotimo o/ 07:48
07:50 lizmat joined
masak waves too 07:50
hoelzro o/
links from breakfast (not sausage links): www.youtube.com/watch?v=WIKqgE4BwAY 07:51
masak plan for the day: help someone else with something. newbies take precedence. :)
moritz I've been mulling an idea for another Perl 6 coding contest 07:52
but learning from masak's mistakes, crowd-source the reviews 07:53
07:53 FROGGS joined
masak ooh 07:53
yeah, doing the reviews all by yourself is fine, until $life intervenes.
crowd-sourcing the reviews might actually work really well if pulled off correctly. 07:54
moritz and then it would still be the one holding the contest (probably me, maybe a second confederate) who selects the winner(s)
07:54 laouji left
nine m: my @a = [$[1, 2, 3], $[4]]; @a>>.say; 07:55
GLRelia rakudo-moar a11e8f: OUTPUT«1␤2␤3␤4␤»
camelia rakudo-moar a7c845: OUTPUT«(Any)␤(Any)␤(Any)␤(Any)␤»
nine m: my @a = [[1, 2, 3], [4]]; @a>>.say;
camelia rakudo-moar a7c845: OUTPUT«4␤3␤1␤2␤»
GLRelia rakudo-moar a11e8f: OUTPUT«1␤2␤3␤4␤»
moritz masak: any ideas on how to pull of the crowed-sourced reviews? 07:56
FROGGS a huperbug /o/
moritz *crowd
nine a plain normal hyper bug
moritz
.oO( cloud-sourced reviews )
m: say () <<+>> ()
GLRelia rakudo-moar a11e8f: OUTPUT«␤»
camelia rakudo-moar a7c845: OUTPUT«␤»
moritz m: say () <<+<< ()
GLRelia rakudo-moar a11e8f: OUTPUT«␤»
camelia rakudo-moar a7c845: OUTPUT«␤»
moritz m: say () >>+<< () 07:57
GLRelia rakudo-moar a11e8f: OUTPUT«␤»
camelia rakudo-moar a7c845: OUTPUT«␤»
moritz didn't I see one of those fail?
maybe fixed in the mean time
laben m: say (1..3) >>+<< (1..4)
GLRelia rakudo-moar a11e8f: OUTPUT«2 4 6 5␤»
camelia rakudo-moar a7c845: OUTPUT«Lists on either side of non-dwimmy hyperop of infix:<+> are not of the same length␤left: 3 elements, right: 4 elements␤ in block <unit> at /tmp/lsJZAZiUg2:1␤␤»
moritz GLR++
laben nine: ^^^ want me to send a PR for this? 07:58
nine laben: please go ahead :)
moritz oh, I mis-read 07:59
nom++
ShimmerFairy moritz: first idea that comes to mind is some sort of anonymous reviewing thing, perhaps some kind of website that presented people a couple of anonymized tests, and likely some guidelines on what's desired.
07:59 [Sno] left, zakharyas left
moritz ShimmerFairy: I'd explicitly not make it anonymous :-) 07:59
07:59 yeahnoob left
dalek ast/glr: e960b57 | (David Warring)++ | / (4 files):
utilize does-ok test assertion
 07:59
ast/glr: 67b2039 | (David Warring [email@hidden.address] | / (4 files):
Merge branch 'does-ok'
ast/glr: 1419d17 | moritz++ | / (4 files):
Merge remote-tracking branch 'origin/master' into glr

Conflicts: S02-types/sigils-and-types.t
moritz fwiw that was merging master into nom in roast 08:00
timotimo ah
moritz ShimmerFairy: allow everybody to submit a review to other people's solutions
does anybody know what's up with the index and rindex tests?
I think something changed in nom, but both nom and the tests have been merged into the glr branches 08:01
ShimmerFairy moritz: sounds good, my idea was just to eliminate knowing anything about the coder beforehand :) . I definitely think the goal for crowd-sourced reviewing would be to have multiple reviews per entry.
moritz ShimmerFairy: aye; and then some kind of summary 08:02
08:03 [Sno] joined 08:04 aborazmeh left, bioexpress left, Sqirrel left 08:05 elohmrow joined
timotimo oh, is this about a p6cc? 08:06
moritz yes 08:07
08:08 RabidGravy joined
ShimmerFairy moritz: I don't know if aggregating reviews would be easier for the contest runner(s) than making the reviews, but I would hope that it is the case :) 08:09
laben nine: github.com/rakudo/rakudo/pull/516
moritz ShimmerFairy: maybe aggregating the reviews can also be crowd-sourced 08:10
nine m: my @a = [$[1, 2, 3], $[4]]; @a>>.say; 08:11
GLRelia rakudo-moar a11e8f: OUTPUT«1␤2␤3␤4␤»
camelia rakudo-moar a7c845: OUTPUT«(Any)␤(Any)␤(Any)␤(Any)␤»
btyler_ waves from SPW 08:12
nine I'm no longer sure this really _is_ a bug. The hyper is supposed to apply the op recursively to the whole list. So .say is called on every item in the nested arrays.
btyler_ need a bigger hackathon room :)
nine TimToady: am I correct?
moritz nine: looks correct to me
and, I mean, where would the Any even come from? 08:13
[$[1, 2, 3], $[4]]
08:13 dakkar joined
FROGGS my expectation was that "my @a = [$[1, 2, 3], $[4]]; @a>>.say;" results in "1 2 3␤4␤" 08:14
masak m: my @a = 1, 2, 3; for @a X @a { .say } 08:15
GLRelia rakudo-moar a11e8f: OUTPUT«1 1␤1 2␤1 3␤2 1␤2 2␤2 3␤3 1␤3 2␤3 3␤»
camelia rakudo-moar a7c845: OUTPUT«1 1␤1 2␤1 3␤2 1␤2 2␤2 3␤3 1␤3 2␤3 3␤»
moritz FROGGS: ... or any permutation thereof
dalek kudo/glr: 8606169 | lizmat++ | src/core/Supply.pm:
Fix some flattening issues with Ranges and emit
nine FROGGS: from S03: Non-structural unary hyper operators produce a hash or array of exactly the same shape as the single argument. The hyper will descend into nested lists and hashes to distribute over the lower-level values just as they distribute over the top-level values that are leaves in the tree. 08:16
[Tux] good morning all
having most of you in Switserland, makes "morning" a well educated guess :)
FROGGS nine: okay, then it seems correct when using .say 08:17
m: my @a = [$[1, 2, 3], $[4]]; my @b = @a>>.++; say @b
GLRelia rakudo-moar a11e8f: OUTPUT«1 2 3 4␤»
camelia rakudo-moar a7c845: OUTPUT«0 0 0 0␤»
[Tux] laben, did you fix anything in glr ot in T::C?
nine m: my @a = [$[1, 2, 3], $[4]]; my @b = @a>>.++; say @b; say @a;
GLRelia rakudo-moar a11e8f: OUTPUT«1 2 3 4␤2 3 4 5␤»
camelia rakudo-moar a7c845: OUTPUT«0 0 0 0␤1 1 1 1␤»
FROGGS \o/ 08:18
m: my @a = [$[1, 2, 3], $[4]]; say @a.perl; my @b = @a>>.++; say @b.perl; say @a.perl;
GLRelia rakudo-moar a11e8f: OUTPUT«[$[1, 2, 3], $[4]]␤[$[1, 2, 3], $[4]]␤[$[2, 3, 4], $[5]]␤»
camelia rakudo-moar a7c845: OUTPUT«[[Any, Any, Any, Any]]<>␤[[0, 0, 0, 0]]<>␤[[1, 1, 1, 1]]<>␤»
FROGGS GLR++ 08:19
laben [Tux]: i did send you some msg to |Tux|. i fixed tests up to 55.t excluded, which needs an input on your part since i am not sure and fixed test 60.t and part of 77.t
[Tux] ah, I have to ssh to the other box to check the log then: [Tux] is at home, |Tux| is at work 08:20
laben you can check irclog
08:21 mr-foobar left 08:22 Sqirrel joined
dalek kudo/glr: 3cd8dd7 | FROGGS++ | src/core/Cool.pm:
fix Cool.{index|rindex} when passing insane offsets
 08:22
08:22 aborazmeh joined, aborazmeh left, aborazmeh joined
FROGGS this ^^^ is also a fix for two tests failing on nom, but I guess we can wait until glr gets merged into nom 08:22
nine laben: I experience some kind of deja vu. Haven't you sent that pull request before? 08:24
laben nine: no, i just posted it on a gist
nine aah, that was it
dalek kudo/glr: fadd434 | laben++ | src/core/metaops.pm:
Fix dwimmy hyper code to correctly consider throw condition
 08:25
kudo/glr: fca2873 | niner++ | src/core/metaops.pm:
Merge pull request #516 from laben/myglr

Fix dwimmy hyper code to correctly consider throw condition
masak m: my @c = <r k d>.pick(*); my @v = <a u o>.pick(*); for ^3 -> $i { print @c[$i]; print @v[$i] }; say ""; 08:26
camelia rakudo-moar a7c845: OUTPUT«rudako␤»
GLRelia rakudo-moar a11e8f: OUTPUT«daroku␤»
masak :D
moritz m: say flat <r k d>.pick(*) Z <a u o>.pick(*) 08:28
GLRelia rakudo-moar a11e8f: OUTPUT«k u d a r o␤»
camelia rakudo-moar a7c845: OUTPUT«d o r u k a␤»
moritz m: say join '', flat <r k d>.pick(*) Z <a u o>.pick(*)
GLRelia rakudo-moar a11e8f: OUTPUT«dakoru␤»
camelia rakudo-moar a7c845: OUTPUT«rodaku␤»
masak moritz++ # functional to my imperative 08:30
m: say join '', <r k d>.pick(*) Z <a u o>.pick(*)
GLRelia rakudo-moar a11e8f: OUTPUT«␤»
camelia rakudo-moar a7c845: OUTPUT«␤»
08:30 larion joined
masak realizes he has to learn to like "flat" from now on :) 08:30
08:31 brrt left
laben m: say (1..3) >>+<< (1..4) 08:31
GLRelia rakudo-moar a11e8f: OUTPUT«2 4 6 5␤»
camelia rakudo-moar a7c845: OUTPUT«Lists on either side of non-dwimmy hyperop of infix:<+> are not of the same length␤left: 3 elements, right: 4 elements␤ in block <unit> at /tmp/ywxB3T6p7x:1␤␤»
laben still not updated
lizmat m: sub a(*@a) { say @a.perl }; a "E"..."A","e"..."a"
GLRelia rakudo-moar a11e8f: OUTPUT«Method 'splice' not found for invocant of class 'List'␤ in block <unit> at /tmp/54J6KyIye9:1␤␤»
camelia rakudo-moar a7c845: OUTPUT«["E", "D", "C", "B", "A", "e", "d", "c", "b", "a"]<>␤»
nine I though GLRelia is updating every 5 minutes?
moritz nine: well, I guess it triggers an updated every 5 minutes, unless a rebuild is currently in progress 08:32
at least that's how I'd do it 08:33
laben the commit hash was not updated, hence i took it as not updated ...
08:35 xfix joined
masak for those listening in from elsewhere: 08:35
we just have a Big Table Refactor, because there were more hackathoners than seats around the table
great fun
laben masak: did you implement a SeatIterator?
btyler_ there's also a spillover hackathon room, if it gets too cozy there 08:36
masak no, we hand-coded everything for maximum speed
FROGGS my spectest hangs :o( 08:37
moritz mine too
after t/spec/rosettacode/sierpinski_triangle.t
FROGGS t/spec/S17-supply/start.t and t/spec/S32-io/IO-Socket-Async.rakudo.moar
lizmat m: sub a(*@a) { say @a.perl }; a 0..10,0...10
GLRelia rakudo-moar a11e8f: OUTPUT«[0, 1, 2, 3, 4, 5, 6, 7, 8, Failure.new(exception => X::Sequence::Deduction.new(from => "9,10,0"), backtrace => Backtrace.new)]␤»
camelia rakudo-moar a7c845: OUTPUT«[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10]<>␤»
tadzik hello hackathon here too! 08:38
FROGGS and we regressed 8 test files :o(
08:38 virtualsue joined
hoelzro masak: and good thing no one got flattened 08:38
moritz huh :(
FROGGS lizmat: did you spectest before&&after your patch? 08:39
I guess I'll re-spectest the last two commits
lizmat I did not spectest, I only tested S17-supply/minmax.t
which now has fewer failing tests 08:40
moritz fwiw I typically capture the spectest output in a file, for example glr-1
lizmat I'll run a spectest now
moritz and then I run grep 'ok$' glr-1 | |cut -d\ -f 1|perl -plE 's/\.rakudo\.moar$/.t/' > t/localtest
FROGGS this is the fallout of the last (two?) commits: gist.github.com/FROGGS/2f06ed42f1755b1235bb
moritz and then a 'make localtest' after a change to make sure there are no regressions 08:41
lizmat so that's good ?
ah no... sorry :-)
itz waves from room 2 08:42
FROGGS lizmat: t/spec/S17-supply/syntax.t, t/spec/S17-supply/unique.t and t/spec/S17-supply/words.t did not fail before
lizmat looks
masak moritz: if you do do a p6cc, maybe I can help with finding tasks. the tasks are very important to get right -- and that's one of the things I believe I got right all three times.
ShimmerFairy except for the commandline magic, that's pretty much how I do it :) (take before and after results, and put in localtest.data any of the interesting files to investigate)
moritz masak: that would be very much appreciated
laben itz: do you have time to check why GLRelia isn't updating anymore?
moritz (the disadvantage of organizing a p6cc is that you can can't participate yourself :-) ) 08:43
dalek rl6-roast-data: ad9443b | coke++ | / (2 files):
another glr run for 'today'
ShimmerFairy masak: task 1: freaking hexslide :P
08:43 virtualsue left
masak ShimmerFairy: is that a variant of ordinary hexslide? :P 08:44
lizmat FROGGS: think I found it
FROGGS lizmat++
08:45 virtualsue joined
[Coke] ^^ that commit also shows a slight regression, if that helps track down versions. 08:46
08:46 Sqirrel left
FROGGS lizmat / others: you can always compare against: github.com/perl6/glr-test-failures...ilures.txt 08:47
it is usually almost up-to-date :o) 08:48
08:48 espadrine left
dalek kudo/glr: 0f43fb6 | lizmat++ | src/core/Supply.pm:
Unbreak some supply tests, break others again :-(
 08:48
ast/glr: 5f85245 | (Stefan Seifert)++ | S17-supply/on.t:
Fix S17-supply/on.t misusing >> when it wants map

  >> is supposed to recurse into nested arrays. It seems like some bug on
nom prevented that. Now that that's fixe on glr, we have to use map instead.
08:50 zakharyas joined
itz laben: yes I've have a look now 08:51
lizmat my @tail := @lol[$m].list.splice(1); # trailing elems of last list added back later # definitely Wrong(TM) 08:52
itz m: say 2+3
GLRelia rakudo-moar a11e8f: OUTPUT«5␤»
camelia rakudo-moar a7c845: OUTPUT«5␤»
masak m: sub foo { my $x = 7; class C { method bar { say $x } } }; C.new.bar
GLRelia rakudo-moar a11e8f: OUTPUT«(Any)␤»
camelia rakudo-moar a7c845: OUTPUT«(Any)␤»
masak should we have a warning when an our-scoped thing sits in a my-scoped thing -- like Perl 5 does? 08:53
FROGGS ahh, that ol' bug :o)
nine m: say [[1..5],[6..10],[11..14]].perl
camelia rakudo-moar a7c845: OUTPUT«[[1, 2, 3, 4, 5], [6, 7, 8, 9, 10], [11, 12, 13, 14]]␤»
GLRelia rakudo-moar a11e8f: OUTPUT«[$[1, 2, 3, 4, 5], $[6, 7, 8, 9, 10], $[11, 12, 13, 14]]␤»
nine m: use Test; is-deeply [[1..5],[6..10],[11..14]], [$[1, 2, 3, 4, 5], $[6, 7, 8, 9, 10], $[11, 12, 13, 14]] 08:54
GLRelia rakudo-moar a11e8f: OUTPUT«ok 1 - ␤»
camelia rakudo-moar a7c845: OUTPUT«not ok 1 - ␤␤# Failed test at /tmp/xgsYOcYtmI line 1␤# expected: [Any, Any, Any, Any, Any, Any, Any, Any, Any, Any, Any, Any, Any, Any]␤# got: [[1, 2, 3, 4, 5], [6, 7, 8, 9, 10], [11, 12, 13, 14]]␤»
08:56 cognominal left
masak does anyone want to pair up on something moderately tricky and semi-secret? 08:57
08:58 GLRelia left
masak on a completely unrelated note, has anyone ever done `take` inside of a grammar? :P 08:59
FROGGS m: my $py = 0 && try { my $py = 42; $py.bla() }
camelia rakudo-moar a7c845: OUTPUT«5===SORRY!5=== Error while compiling /tmp/GaBH5Z1ixo␤Cannot use variable $py in declaration to initialize itself␤at /tmp/GaBH5Z1ixo:1␤------> 3my $py = 0 && try { my $py = 42; $7⏏5py.bla() }␤ expecting any of:␤ term␤»
FROGGS m: my $py = 0 && try { { my $py = 42; $py.bla()}() } 09:00
camelia ( no output )
moritz should probably start the nom release process
09:01 abraxxa joined
itz Makefile:422: recipe for target 'CORE.setting.moarvm' failed 09:02
FROGGS itz: does it reconfigure on each build? 09:03
itz maybe cruft left in my env unless others can't compile GLR
FROGGS because lately a new file was added to the setting build (SomethingIterator.pm or so)
itz FROGGS: ah thanks 09:04
FROGGS [ptc]: it is now rakudobugged as RT #125921 09:07
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=125921
09:07 brrt joined
masak ++moritz # nom release process 09:08
by the way, colomon++ has now joined the hackathon too! \o/ 09:09
09:09 colomon joined
colomon hello hackathon! 09:09
masak colomon! \o/
FROGGS \o\
masak starts and looks over his shoulder
aie!
colomon: don't sneak up on me like that! 09:10
:P
FROGGS hoelzro: this is ven in case you've not seen him yet
dalek kudo/glr: 04a8f10 | (Stefan Seifert)++ | src/core/Supply.pm:
Fix Supply.batch flattening the results too much

batch is supposed to pass an array to emit. Instead it passed a list of arguments due to the single arg rule.
masak ooh, and now vendethiel arrived! \o/
hoelzro ahhhhh
FROGGS ahh, he has a badge... nvm :o)
masak I hope you all realize... 09:11
...me and vendethiel in the same room...
...there will be macros. :)
[ptc] FROGGS: sweet, thanks
yoleaux 20 Aug 2015 16:27Z <moritz> [ptc]: a perl6-examples job on hack was running for 72 CPU hours before I killed it; please investigate (and add a CPU limit to all cron jobs)
09:11 espadrine joined
yoleaux 21 Aug 2015 14:42Z <moritz> [ptc]: I've set some ulimits in perl6-examples' update-and-sync, and updated the checkout on hack; that was in a detached HEAD previously. If master sounds wrong to you, please use "git reflog" to restore the old state (and fix the issue otherwise) 09:11
colomon NOOOOOOOO!
yoleaux 25 Aug 2015 10:33Z <_itz> [ptc]: examples.perl6.org appears to have stopped updating
masak colomon: mwhahaha
ShimmerFairy
.oO(rows of apple comps)
 09:13
[ptc] moritz: sorry about the perl6-examples jobs! I've been only a lurker for a while ($work has been pretty busy) and hadn't seen your message
tadzik and sorted thinkpads :P
nine masak: yeah we hope so :)
[ptc] itz: thanks for the message. Will look into it over the next couple of days :-)
itz [ptc]: are in .ch .. I was planning on duplicating the examples.perl6.org update process elsewhere as a dev instance? 09:14
[ptc] waves from the next room
itz ah right
lizmat m: sub a(*@a) { say @a.perl }; a 0..10,0...10
camelia rakudo-moar a7c845: OUTPUT«[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10]<>␤»
lizmat m: sub a(*@a) { say @a.perl }; a 0..10,0...10
camelia rakudo-moar a7c845: OUTPUT«[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10]<>␤»
lizmat m: sub a(*@a) { say @a.perl }; a 0...10 09:15
camelia rakudo-moar a7c845: OUTPUT«[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10]<>␤»
lizmat m: sub a(*@a) { say @a.perl }; a 0..10
camelia rakudo-moar a7c845: OUTPUT«[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10]<>␤»
lizmat m: sub a(*@a) { say @a.perl }; a 0..5,6...10
camelia rakudo-moar a7c845: OUTPUT«[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10]<>␤»
lizmat m: sub a(*@a) { say @a.perl }; a 0..5,0...10
camelia rakudo-moar a7c845: OUTPUT«Unable to deduce arithmetic or geometric sequence from 4,5,0 (or did you really mean '..'?)␤ in sub a at /tmp/aIraUSs47h:1␤ in block <unit> at /tmp/aIraUSs47h:1␤␤Actually thrown at:␤ in sub a at /tmp/aIraUSs47h:1␤ in block <unit> at /tmp/aI…»
colomon so what’s the plan of attack here? should I grab a random hunk of failing roast-glr and try to make it work? should I ignore glr and try to improve the smoker? fix modules? 09:17
masak colomon: I think people would be happy if you jumped into GLR. 09:19
the other stuff dosuns nice too, though.
colomon thinks he will have more time to hack on p6 during this hackathon than he did in the entire rest of the year so far... 09:20
nine colomon: yeah, just grab a failing test :)
Maybe we should start coordinating a bit on the channel though with so many people picking tests and failures becoming fewer and fewer 09:21
dalek ast/glr: 18f810a | lizmat++ | S17-supply/minmax.t:
Fix test by not using sequences
colomon sounds reasonable
nine Only 81 failing now!
colomon nine: running a spectest already
lizmat that's 80 now!
09:22 GLRelia joined
itz m: say 2+3 09:22
camelia rakudo-moar a7c845: OUTPUT«5␤»
GLRelia rakudo-moar a11e8f: OUTPUT«5␤»
masak yay, I just helped btyler_++ fix his code! \o/
09:22 donaldh joined
masak feels warm an fuzzy inside 09:23
FROGGS .oO( Fix test by avoiding code that exposes the error )
lizmat well, the error is with sequences, not with supplies
FROGGS lizmat: are there other tests for that reverse() error?
lizmat it's not reverse() that is a problem, it's sequences
m: say 1...10,1...10
camelia rakudo-moar a7c845: OUTPUT«1 2 3 4 5 6 7 8 9 10␤»
GLRelia rakudo-moar a11e8f: OUTPUT«Method 'splice' not found for invocant of class 'List'␤ in block <unit> at /tmp/EyF2HOltok:1␤␤»
FROGGS lizmat: are there tests for that? 09:24
nine m: say 1..10,1...10
camelia rakudo-moar a7c845: OUTPUT«1 2 3 4 5 6 7 8 9 10␤»
GLRelia rakudo-moar a11e8f: OUTPUT«1 2 3 4 5 6 7 8 9 10␤»
nine m: say 1..10...10
camelia rakudo-moar a7c845: OUTPUT«1 2 3 4 5 6 7 8 9 10␤»
GLRelia rakudo-moar a11e8f: OUTPUT«1 2 3 4 5 6 7 8 9 10␤»
nine m: say 1..10,0...10
camelia rakudo-moar a7c845: OUTPUT«1 2 3 4 5 6 7 8 9 10␤»
GLRelia rakudo-moar a11e8f: OUTPUT«1 2 3 4 5 6 7 8 9 10␤»
laben m: dd 1...10,1...10 09:25
camelia rakudo-moar a7c845: OUTPUT«(1, 2, 3, 4, 5, 6, 7, 8, 9, 10)␤»
GLRelia rakudo-moar a11e8f: OUTPUT«Method 'splice' not found for invocant of class 'List'␤ in block <unit> at /tmp/v5UaX9Eq09:1␤␤»
laben m: dd 1..10,1...10
camelia rakudo-moar a7c845: OUTPUT«(1, 2, 3, 4, 5, 6, 7, 8, 9, 10)␤»
GLRelia rakudo-moar 11e8f: OUTPUT«(1, 2, 3, 4, 5, 6, 7, 8, 9, 10)␤»
laben m: dd ,1...10
camelia rakudo-moar a7c845: OUTPUT«5===SORRY!5=== Error while compiling /tmp/dwW79VhrtQ␤Preceding context expects a term, but found infix , instead␤at /tmp/dwW79VhrtQ:1␤------> 3dd ,7⏏051...10␤»
GLRelia rakudo-moar 11e8f: OUTPUT«5===SORRY!5=== Error while compiling /tmp/GuUZkO5nXx␤Preceding context expects a term, but found infix , instead␤at /tmp/GuUZkO5nXx:1␤------> 3dd ,7⏏051...10␤»
laben m: dd 1...10,
camelia rakudo-moar a7c845: OUTPUT«(1, 2, 3, 4, 5, 6, 7, 8, 9, 10)␤»
GLRelia rakudo-moar 11e8f: OUTPUT«(1, 2, 3, 4, 5, 6, 7, 8, 9, 10)␤»
laben m: dd 1...10
camelia rakudo-moar a7c845: OUTPUT«(1, 2, 3, 4, 5, 6, 7, 8, 9, 10)␤» 09:26
GLRelia rakudo-moar 11e8f: OUTPUT«(1, 2, 3, 4, 5, 6, 7, 8, 9, 10)␤»
tadzik ..wth is "Bytecode stream version too low"? :O
TimToady m: say (1,2,3 ... 2)
camelia rakudo-moar a7c845: OUTPUT«1 2␤»
GLRelia rakudo-moar 11e8f: OUTPUT«1 2␤»
TimToady m: my @x; @x[0,2,4...$end]
camelia rakudo-moar a7c845: OUTPUT«5===SORRY!5=== Error while compiling /tmp/_SokNrE8UA␤Variable '$end' is not declared. Did you mean '&end'?␤at /tmp/_SokNrE8UA:1␤------> 3my @x; @x[0,2,4...7⏏5$end]␤»
GLRelia rakudo-moar 11e8f: OUTPUT«5===SORRY!5=== Error while compiling /tmp/GIGORVrSkg␤Variable '$end' is not declared. Did you mean '&end'?␤at /tmp/GIGORVrSkg:1␤------> 3my @x; @x[0,2,4...7⏏5$end]␤» 09:27
jnthn ohhai o/
jnthn is at l'hackathon now also :)
lizmat jnthn o/
itz the GLRelia git sha should correct shortly
colomon is going to look at S32-list/pick.t 09:28
itz m: say "sha hash"
camelia rakudo-moar a7c845: OUTPUT«sha hash␤»
GLRelia rakudo-moar 04a8f1: OUTPUT«sha hash␤»
tadzik FROGGS: github.com/tadzik/panda/issues/178 is there a reason bundle() shouldn't create source-url instead?
dalek kudo/glr: 30161a5 | FROGGS++ | src/Perl6/Grammar.nqp:
reset to-be-initialied variable name in blockoid

Before we did that in pblock, which means that e.g. blorsts did not get handled correctly. Fixes RT #125921.
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=125921
masak m: say (flat 0..9, 'a'..'f').roll(40).join 09:29
camelia rakudo-moar a7c845: OUTPUT«30a840c3c9475c53db38dcd25d934dd6ffc6afb8␤»
GLRelia rakudo-moar 04a8f1: OUTPUT«fd407e3a43cb55d696235b75b5b39189fde0af01␤»
laben m: say (1..3) >>+<< (1..4)
camelia rakudo-moar a7c845: OUTPUT«Lists on either side of non-dwimmy hyperop of infix:<+> are not of the same length␤left: 3 elements, right: 4 elements␤ in block <unit> at /tmp/sz3CKfl93d:1␤␤»
GLRelia rakudo-moar 04a8f1: OUTPUT«Lists on either side of non-dwimmy hyperop of infix:<+> are not of the same length␤left: 3 elements, right: 4 elements␤ in block <unit> at /tmp/jhRdFT93QM:1␤␤»
09:29 domidumont left
laben finally ^^^ 09:29
09:29 cognominal joined
FROGGS tadzik: I guess that is just an oversight, though, please also look at S22, perhaps it is called 'source' there 09:29
dalek osystem: f720f31 | tadzik++ | META.list:
Retire MuEvent
09:30 domidumont joined
tadzik FROGGS: ah, it's support.source in S22, and panda just relies on non-S22 field here 09:31
S22 doesn't seem to know about "here's where you probably downloaded this from" :)
and it really only makes sense for the makeshit CPAN we created
FROGGS tadzik: hmmm... can we make this sane and e.g. let panda understand both?
tadzik sounds good
09:32 CIAvash joined
colomon m: say @a.pick(1).WHAT 09:33
camelia rakudo-moar a7c845: OUTPUT«5===SORRY!5=== Error while compiling /tmp/A55Kbj_K3M␤Variable '@a' is not declared␤at /tmp/A55Kbj_K3M:1␤------> 3say 7⏏5@a.pick(1).WHAT␤»
GLRelia rakudo-moar 04a8f1: OUTPUT«5===SORRY!5=== Error while compiling /tmp/cNRQL5qAxP␤Variable '@a' is not declared␤at /tmp/cNRQL5qAxP:1␤------> 3say 7⏏5@a.pick(1).WHAT␤»
colomon m: say (1..100).pick(1).WHAT
camelia rakudo-moar a7c845: OUTPUT«(List)␤»
GLRelia rakudo-moar 04a8f1: OUTPUT«(Seq)␤»
colomon Seq is correct for GLR, no?
jnthn colomon: yeah, correct 09:34
colomon okay, patching roast-glr
09:34 Ven joined
masak Ven! \o/ 09:34
Ven \o/, spw2015 ! 09:35
colomon \o/ 09:36
dalek ast/glr: 3bddd39 | colomon++ | S32-list/pick.t:
pick returns a Seq now.
colomon okay, hmmm … S32-list/classify.t ?
[Coke] m: say ~3.14159265358979323846264338327950288419716939937510 09:38
camelia rakudo-moar a7c845: OUTPUT«3.14159265358979338245148937858175486326217651367188␤»
GLRelia rakudo-moar 30161a: OUTPUT«3.14159265358979338245148937858175486326217651367188␤»
colomon m: say classify( {.comb}, 100 .. 119, 104, 119 ).perl
camelia rakudo-moar a7c845: OUTPUT«Hash[Any,Any].new("1" => Hash[Any,Any].new("0" => Hash[Any,Any].new("0" => [100], "1" => [101], "2" => [102], "3" => [103], "4" => [104, 104], "5" => [105], "6" => [106], "7" => [107], "8" => [108], "9" => [109]), "1" => Hash[Any,Any].new("0" => [110], "1"…»
GLRelia rakudo-moar 30161a: OUTPUT«Hash[Any,Any].new()␤»
colomon that looks like a very real bug
dalek ast/glr: c86bb56 | FROGGS++ | S04-declarations/my.t:
add test for RT #125371
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=125371
09:39 Sqirrel joined
FROGGS [ptc]: it is fixed in rakudo/glr now 09:39
[Coke] m: say ~3.14159265358979323
camelia rakudo-moar a7c845: OUTPUT«3.14159265358979323␤»
GLRelia rakudo-moar 30161a: OUTPUT«3.14159265358979323␤»
timotimo ohai colomon, you're the gentleman in the corner with the apple laptop? 09:40
colomon indeed 09:41
timotimo yay
dalek kudo/glr: 8f5dd0d | lizmat++ | src/core/io_operators.pm:
Remove say(\x) / note(\x) candidates

These seem like premature optimizations in the GLR, and they break behaviour. They're just commented out for future reference / complete removal when we feel certain that's the way to go
itz "with the apple laptop" is hardly a unique identifier :)
colomon itz: sitting in the corner is
;)
nine colomon: we do have a free seat at the table here
TimToady not sure how many gentlemen there are here either...
09:42 aborazmeh left
colomon nine: right now I’m enjoying the breeze from the window I’m next to. ;) 09:42
nine colomon: good point :) 09:43
nine is looking at t/spec/S17-supply/stable.t
timotimo jnthn: i have an early implementation of nfa_to_statelist; i'd like you to take a look at it just to verify that what i'm doing with the MVMROOTs and such is sane and not totally wrong; would you like a diff?
tadzik moritz: ecocystem-api's projects.json actually generated from github.com/perl6/ecosystem/blob/ma...telist.pl? 09:44
moritz tadzik: yes
tadzik ok, cool 09:45
colomon hmmm… classify-list is calling loads of nqp routines.
tadzik the "New run at" line from update.log made me wonder
colomon m: say List ~~ Seq
camelia rakudo-moar a7c845: OUTPUT«5===SORRY!5=== Error while compiling /tmp/JPpt_1Jn0f␤Undeclared name:␤ Seq used at line 1. Did you mean 'Set'?␤␤» 09:46
GLRelia rakudo-moar 8f5dd0: OUTPUT«False␤»
colomon m: say Seq ~~ List
camelia rakudo-moar a7c845: OUTPUT«5===SORRY!5=== Error while compiling /tmp/e9iLq7g7lh␤Expected a term, but found either infix ~~ or redundant prefix ~␤ (to suppress this message, please use a space like ~ ~)␤at /tmp/e9iLq7g7lh:1␤------> 3say Seq ~~7⏏5 List␤»
GLRelia rakudo-moar 8f5dd0: OUTPUT«False␤»
timotimo gist.github.com/timo/04a201f9c5fe9b5bd752 ← jnthn, this is the diff
moritz colomon: both are Iterable
colomon moritz: Aha! That may be what we need here
m: say “here”.comb.WHAT 09:47
camelia rakudo-moar a7c845: OUTPUT«(List)␤»
GLRelia rakudo-moar 8f5dd0: OUTPUT«Invocant requires an instance of type Seq, but a type object was passed. Did you forget a .new?␤ in block <unit> at /tmp/yHFdJf4qYi:1␤␤»
colomon ???
09:47 araujo left
timotimo er, huh? :) 09:47
colomon m: say “here”.comb ~~ Iterable
camelia rakudo-moar a7c845: OUTPUT«True␤»
GLRelia rakudo-moar 8f5dd0: OUTPUT«True␤»
colomon m: say “here”.comb ~~ Seq
camelia rakudo-moar a7c845: OUTPUT«5===SORRY!5=== Error while compiling /tmp/w7DGW_RUBf␤Undeclared name:␤ Seq used at line 1. Did you mean 'Set'?␤␤»
GLRelia rakudo-moar 8f5dd0: OUTPUT«True␤»
colomon m: say “here”.comb ~~ List 09:48
camelia rakudo-moar a7c845: OUTPUT«True␤»
GLRelia rakudo-moar 8f5dd0: OUTPUT«False␤»
colomon m: say nqp::istype(“here”.comb, Iterable)
camelia rakudo-moar a7c845: OUTPUT«===============================================================================␤The use of nqp::operations has been deprecated for non-CORE code. Please␤change your code to not use these non-portable functions. If you really want␤to keep using nqp:…»
GLRelia rakudo-moar 8f5dd0: OUTPUT«===============================================================================␤The use of nqp::operations has been deprecated for non-CORE code. Please␤change your code to not use these non-portable functions. If you really want␤to keep using nqp:…»
colomon bother
09:49 rindolf joined, llfourn left
colomon will nqp::istype work for subtypes? 09:49
dalek kudo/nom: eab03c3 | coke++ | src/core/ (2 files):
RT #123760 - add Str matchers for comb
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=123760
timotimo you just have to "use nqp" in front, colomon
that'll shut it up
laben m: use nqp; say nqp::istype(“here”.comb, Iterable)
dalek ast: 3871847 | coke++ | S32-str/comb.t:
RT #123760 update/add tests for Str-based comb
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=123760
camelia rakudo-moar a7c845: OUTPUT«1␤»
GLRelia rakudo-moar 8f5dd0: OUTPUT«1␤»
colomon timotimo++
timotimo sadly, it puts the "…" right in front of that hint
because of all those "======"
colomon timotimo: extra sadly, when I got the same message on my local perl6, I didn’t read the second half (with use nqp) because the first half of the message was identical to what I got on GLRelia. 09:51
laben *elia runs on RESTRICTED, no? we could reduce the number of === in that
09:51 andreoss joined, sno joined
timotimo heh. :) 09:51
09:51 [Sno] left
timotimo that seems more like a hack than an elegant solution 09:51
we could just have a CAMELIA.setting instead :) 09:52
colomon so, my first rakudo-glr patch coming up in a min. (fixed classify, timotimo++)
timotimo very cool!
colomon errr, more than a minute, spectesting is slow.
laben make number of === configurable at runtime/builtime?
dalek osystem: 058a95d | tadzik++ | META.list:
Retire pandabrew
laben i think having that at runtime with a flag or a $*VAR could be the best solution 09:53
[Coke] r: constant @t = [\+] 1..5; say (0,@t Z^.. @t).perl;
GLRelia rakudo-moar 8f5dd0: OUTPUT«This Seq has already been iterated, and its values consumed␤ in block <unit> at /tmp/tmpfile:1␤␤»
..rakudo-jvm 8f5dd0: OUTPUT«Can't open perl script "/home/steve/rakudo-inst/bin/eval-client.pl": No such file or directory␤»
camelia rakudo-{moar,jvm} a7c845: OUTPUT«(1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15)␤»
ShimmerFairy I think that's actually a hard-coded ===== string for the nqp message, IIRC.
09:53 andreoss left, andreoss joined
colomon JOBS=4? 09:53
09:54 andreoss left
dalek c: ba14f7b | paultcochrane++ | htmlify.p6:
Work around rakudobug RT #125921

  FROGGS++ for adding this to rakudobug. This is also now fixed in
rakudo/glr, so we can change the code back if desired after the glr merge.		 09:55
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=125921
colomon will I get extra spectest failures because of TEST_JOBS=4 ? 09:56
timotimo usually not 09:57
colomon \o/
I must say, running with TEST_JOBS=4 is making my laptop heat up…. :)
timotimo but nine has found a case where it does break a test on his machine
yeah, i can imagine 09:58
[Coke] started adding [GLR] to tickets that are/should be/maybe GLR related. - rt.perl.org/Ticket/Display.html?id=113870
09:58 Ven left
timotimo my laptop has a derf where it odesn't properly turn up the fan speed 09:58
ShimmerFairy It tends to break lock.t for me, in S17
timotimo so when i'm alone i just tend to "echo 'level disengaged' | sude tee fan"
("level 7" would be the highest speed the fan can turn on via regulation, but "level disengaged" just turns off the regulator and it just goes another 2x faster than "fastest speed") 09:59
09:59 araujo joined
[Coke] m: <a a>.squish(:with(-> $a, $b {1})).perl.say 10:00
camelia rakudo-moar a7c845: OUTPUT«()␤»
GLRelia rakudo-moar 8f5dd0: OUTPUT«().Seq␤»
moritz uhm, why doesn't that return ('a') or ('a').Seq ? 10:01
colomon wouldn’t be surprised if squish is broken in the same way classify was (well, is, as I haven’t pushed yet) 10:02
dalek pan style="color: #395be5">perl6-examples: 5e42f33 | (Steve Mynott)++ | categories/cookbook/16processes/16-14-send-signal.pl:
simple send signal example using nativecall
 10:03
moritz colomon: push, push, push! (www.youtube.com/watch?v=ICTmiUhea3g :-) 10:04
colomon moritz: I’m spectesting first, and that’s slow
timotimo maybe we should have brought a very beefy machine to put into some corner where no ears are close by to do all our spec testing for us
moritz colomon: I tend to spectest my GLR changes on hack, because it's faster than my laptop :-) 10:05
timotimo like a 32-core machine local to the venue
10:05 Sqirrel left
colomon timotimo, moritz: yeah, I’m spinning on spectests on my big box back in Michigan now… ;) 10:05
my TEST_JOBS=4 run on OS X seems like it might be wedged on t/spec/S17-procasync/no-runaway-file-limit.t ?
10:07 brrt left
FROGGS m: sub foo(:xen(:$xin)) { say $xin }; foo xin => 23 10:08
camelia rakudo-moar a7c845: OUTPUT«23␤»
GLRelia rakudo-moar 8f5dd0: OUTPUT«23␤»
dalek pan style="color: #395be5">perl6-examples: 7c1187a | (Sue Spence)++ | categories/cookbook/01strings/01-00introduction.pl:
Add examples for $obj.say. Fix minor typo.
pan style="color: #395be5">perl6-examples: db65979 | (Sue Spence)++ | categories/cookbook/01strings/01-01substrings.pl:
Pugs -> Rakudo
pan style="color: #395be5">perl6-examples: 234f01b | (Sue Spence)++ | categories/cookbook/01strings/01-04converting-values.pl:
Answered some commenty questions, added a unicode char
pan style="color: #395be5">perl6-examples: 2049f7e | (Steve Mynott)++ | categories/cookbook/01strings/01-0 (3 files):
Merge pull request #27 from virtualsue/master

Minor updates
10:09 aborazmeh joined, aborazmeh left, aborazmeh joined
[ptc] itz: you beat me to it :-) 10:09
masak TimToady, jnthn: Ven and I think that maybe we are too attached to push's current semantics. 10:10
10:10 Sqirrel joined
masak and that it's not so unreasonable to do `@a.push(|@b)` 10:10
tadzik masak: I reckon you'll like that: github.com/tadzik/panda/commit/cfe6ccb8c54 :) 10:11
masak (not saying we're right, just that it's something worth considering)
tadzik: yaaaaay
[Coke] m: say ~('Y'..'z')
camelia rakudo-moar a7c845: OUTPUT«Y Z [ \ ] ^ _ ` a b c d e f g h i j k l m n o p q r s t u v w x y z␤»
GLRelia rakudo-moar 8f5dd0: OUTPUT«Y Z [ \ ] ^ _ ` a b c d e f g h i j k l m n o p q r s t u v w x y z␤»
itz [ptc]: she was sitting opposite :)
ShimmerFairy masak: Looking at it, it feels natural at least, considering the original meaning of prefix:<|> :)
tadzik it's not even *that* drastically slow
loading HTTP::UA goes from 0.8 secs to 5.4 on my machine
[Coke] m: say ~('Y'..'_') 10:13
camelia rakudo-moar a7c845: OUTPUT«Y Z [ \ ] ^ _␤»
GLRelia rakudo-moar 8f5dd0: OUTPUT«Y Z [ \ ] ^ _␤»
[Coke] m: say ~('Y'..'AB')
camelia rakudo-moar a7c845: OUTPUT«␤»
GLRelia rakudo-moar 8f5dd0: OUTPUT«␤»
jnthn masak: |@b is going to suck if the thing is 100s of items long 10:14
dalek kudo/glr: ca6b9e6 | colomon++ | src/core/Hash.pm:
Switch List to Iterable.

Pretty sure this was what was originally meant, and Iterable treats Seq in the same fashion. Fixes the spectest fail in S32-list/classify.t.
colomon never did get a useful spectest run, but is pretty sure the patch is correct.
jnthn masak: Of note, on JVM it'll actually break if you make the thing more than about 255 items long.
10:15 Ven joined
jnthn masak: Because that is the maximum number of args you can pass on the JVM 10:15
nine colomon: +1
dalek pan style="color: #395be5">perl6-examples: 8dd4a70 | (Steve Mynott)++ | categories/cookbook/16processes/16-18-catch-control-c.pl:
Catch ctrl-c to ignore or execute code example
pan style="color: #395be5">perl6-examples: 8f84888 | (Steve Mynott)++ | categories/cookbook/01strings/01-0 (3 files):
Merge branch 'master' of github.com:perl6/perl6-examples
ast: 646e60f | coke++ | S03-operators/range.t:
RT #124542 #124543 - remove/update dodgy tests.
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=124542
nine colomon: similar bugs may be in Baggy.pm 10:16
colomon nine: I will investigate
nine :)
Haven't really looked at Bag/Mix/... bugs, because I don't yet know what they are :) 10:17
10:17 ab6tract joined
colomon nine++: looks like exactly the same code that I just changed, good call 10:17
not sure if this is triggering a spectest failure, which suggests we need a new test first... 10:18
[Coke] Down to 1.042K tickets.
nine colomon: t/spec/S02-types/bag.rakudo.moar and t/spec/S02-types/baghash.rakudo.moar are failing right now
colomon: maybe your fix helps there
colomon oooo!
nine++
nine colomon++ # fixing things :) 10:19
colomon nine: it feels really good to be digging back into p6 in a useful way. :)
10:19 jkva joined
FROGGS m: say <foo bar baz> Z=> 1 xx * 10:19
camelia rakudo-moar a7c845: OUTPUT«foo => 1 bar => 1 baz => 1␤»
GLRelia rakudo-moar 8f5dd0: OUTPUT«foo => 1bar => 1baz => 1␤»
FROGGS m: say (<foo bar baz> Z=> 1 xx *).perl
camelia rakudo-moar a7c845: OUTPUT«(:foo(1), :bar(1), :baz(1))␤»
GLRelia rakudo-moar 8f5dd0: OUTPUT«(:foo(1), :bar(1), :baz(1)).Seq␤»
timotimo nqp-m: say NQPList
camelia nqp-moarvm: OUTPUT«Confused at line 2, near "say NQPLis"␤ at gen/moar/stage2/NQPHLL.nqp:516 (/home/camelia/rakudo-inst-2/share/nqp/lib/NQPHLL.moarvm:panic:105)␤ from gen/moar/stage2/NQP.nqp:921 (/home/camelia/rakudo-inst-2/share/nqp/lib/nqp.moarvm:comp_unit:872)␤ from gen/…»
timotimo nqp-m: say(NQPList)
camelia nqp-moarvm: OUTPUT«␤»
10:19 virtualsue left
timotimo nqp-m: say(NQPMu) 10:20
camelia nqp-moarvm: OUTPUT«␤»
10:20 virtualsue joined
lizmat m: say "foo".rindex("f",1e35) # FROGGS, shouldn't that die now ? 10:20
FROGGS m: my %h = <foo bar baz> Z=> 1 xx *; for %h.keys { .say }
camelia rakudo-moar a7c845: OUTPUT«Nil␤»
GLRelia rakudo-moar 8f5dd0: OUTPUT«0␤»
camelia rakudo-moar a7c845: OUTPUT«bar␤baz␤foo␤»
GLRelia rakudo-moar 8f5dd0: OUTPUT«bar␤baz␤foo␤»
timotimo ah, needs to be NQPArray
nqp-m: say(NQPArray)
camelia nqp-moarvm: OUTPUT«␤»
timotimo that doesn't tell me anything, though
FROGGS lizmat: I don't think so
lizmat: the offset is longer than the string, so you get Nil... err, supposed to get Nil 10:21
colomon nine: hmm, I think the /S02-types/bag.t are likely to be glitches in the bag constructor. I will look at these once I look at classify. :)
masak FROGGS: Ven and I are discussing gist.github.com/masak/3438222 now
FROGGS masak: ool 10:22
cool*
lizmat FROGGS: ok
FROGGS lizmat: so it is a bug that it does not return Nil 10:23
lizmat yup
tadzik masak: I feel like both are just avoiding the need of a C parser that'll automatically generate this stuff :)
FROGGS m: use nqp; say nqp::unbox_i 1e35.Int
camelia rakudo-moar eab03c: OUTPUT«0␤»
GLRelia rakudo-moar ca6b9e: OUTPUT«0␤»
FROGGS :o(
10:23 rurban left
lizmat yeah :-( 10:23
jnthn FROGGS: Hm, didn't you commit the check for that overflow? 10:24
FROGGS potentially it fits into an MVMuint64, and will later be casted into an MVMint64
lizmat FROGGS: so *that*'s the bug, right ?
FROGGS lizmat: aye
jnthn Oh, 1e35 easily fits, no?
colomon lizmat: do I trigger the Baggy classify-list routine by using into=>Bag.new when I call classify?
[Coke] Is this related to the failure on os x about the overflow check?
jnthn Oh, maybe not :)
10:24 muraiki joined
lizmat colomon: without looking at the code, I think it does 10:25
FROGGS m: say 1e35.Int.fmt('%b').chars
camelia rakudo-moar eab03c: OUTPUT«117␤»
GLRelia rakudo-moar ca6b9e: OUTPUT«117␤»
FROGGS m: say 1e35.Int.fmt('%b')
camelia rakudo-moar eab03c: OUTPUT«100110100001001100001011100101100011101001101100000100000000000000000000000000000000000000000000000000000000000000000␤»
GLRelia rakudo-moar ca6b9e: OUTPUT«100110100001001100001011100101100011101001101100000100000000000000000000000000000000000000000000000000000000000000000␤»
[Coke] t/spec/S02-types/declare.t - fails test #19
lizmat [Coke]: is that a new fail ?
FROGGS m: use nqp; say nqp::unbox_i 1e36.Int
camelia rakudo-moar eab03c: OUTPUT«0␤»
GLRelia rakudo-moar ca6b9e: OUTPUT«0␤»
FROGGS m: use nqp; say nqp::unbox_i 1e34.Int
camelia rakudo-moar eab03c: OUTPUT«3458764513820540928␤»
GLRelia rakudo-moar ca6b9e: OUTPUT«3458764513820540928␤»
FROGGS m: use nqp; say nqp::unbox_i (-1e35).Int 10:26
camelia rakudo-moar eab03c: OUTPUT«0␤»
GLRelia rakudo-moar ca6b9e: OUTPUT«0␤»
FROGGS ??
[Coke] been failing since the test was added.
colomon lizmat: "... a large array goes into a bar - I mean bag - with 50k elems and lives" :)
FROGGS this was working just the other day
$ perl6 -e 'use nqp; say nqp::unbox_i 1e36.Int'
Cannot unbox 120 bit wide bigint into native integer
[Coke] I mentioned it previously in backlog, but haven't opened a ticket, my bad.
FROGGS maybe we are on an old moarvm?
lizmat $ 6 'use nqp; say nqp::unbox_i 1e36.Int' 10:27
0
that's the latest rakudo on GLR and OS X
timotimo jnthn: how can i be segfaulting on gist.github.com/timo/04a201f9c5fe9...1-txt-L234 - this line? :o
FROGGS I'll bump moar revision for nqp+glr
dalek pan style="color: #395be5">perl6-examples: 437a6c4 | paultcochrane++ | categories/euler/prob (4 files):
Purge hard tabs in examples
 10:28
p: 02910ff | FROGGS++ | tools/build/MOAR_REVISION:
bump moar revision
kudo/glr: 23b3d29 | FROGGS++ | tools/build/NQP_REVISION:
bump nqp/moar revision
 10:29
FROGGS lizmat / [Coke]: this should do I hope
JimmyZ timotimo: thre is a MVMROOT, which cause the line is wrong.
|Tux| can I see from perl6 --version if it is nom or glr?
yoleaux 26 Aug 2015 17:43Z <laben> |Tux|: i fixed up to test file 50_utf8.t, i'm now at 55_combi.t where the problem is trying to assign a list of chars (instead of Str) to both $!esc and $!quo. what did you mean to do?
26 Aug 2015 21:27Z <laben> |Tux|: i fixed 60_samples.t after skipping 55_combi.t, i got up to 90/114 tests of 77_getall.t passing, still needs a lot of work
lizmat FROGGS: just pulled, building now 10:30
timotimo JimmyZ: so in reality the problem must be on a different line?
so ... how do i find the right line? turn on disassembly? X_x
JimmyZ timotimo: yeah, change it to the function ...
FROGGS |Tux|: not reliably I guess
timotimo you mean MVM_gc_temp_roots_push or what it's called? 10:31
JimmyZ yes
dalek pan style="color: #395be5">perl6-examples: 6e50a73 | paultcochrane++ | bin/ (2 files):
Remove now outdated utility scripts

Adding a collaborator can happen either via the GitHub website, or if it has to be scripted, via the GitHub api. The twitter2irc script doesn't seem to be used at all.
[Coke] FROGGS: my failure is on nom.
FROGGS [Coke]: I can bump nom too, hold on
10:31 virtualsue left
timotimo i'll have a look 10:32
JimmyZ timotimo: or use MSVC, which seems to be a little friendly to macro
timotimo: jnthn may know :P
dalek kudo/nom: a46b09c | FROGGS++ | tools/build/NQP_REVISION:
bump nqp/moar revision
 10:33
FROGGS [Coke]: ^^
JimmyZ anyone could merge github.com/rakudo/rakudo/pull/506? 10:34
t is safe now
[Coke] 1) build rakudo 2) install it. 3) panda rebootstrap 4) install linenoise ... 5) build directory's ./perl6-m now supports linenoise? 10:36
timotimo JimmyZ: now the segfault comes from the next MVMROOT …
JimmyZ timotimo: you need changed all the MVMROOT of this file, as far as I know
[Coke] ah, yes, because the ./perl6-m handler is using the installed prefix as a libpath.
timotimo $2 = {fates = 0xc30180, num_states = 0, num_state_edges = 0x38000000000001, states = 0x6075b8} 10:37
that looks sane!
no problems here!
Ulti /win 2 10:38
erk
colomon cannot find any sign that roast actually tests Baggy.classify-list (even indirectly)
[ptc] itz: found lots of examples-related processes were still running from 22nd of August. Have just killed them all. Running update-and-sync by hand should hopefully work again now.
moritz colomon: then that should be corrected :-) 10:39
colomon moritz: I’m working on it. but it’s tricky, I need to figure out what it is supposed to do. ;)
timotimo grabs one of these bites
moritz colomon: :-)
colomon aha!
jnthn timotimo: Is the debugger pointing at the wrong line 'cus it's confused by the macro? 10:40
colomon m: say (1..10).classify(* %% 2) 10:42
camelia rakudo-moar eab03c: OUTPUT«False => 1 3 5 7 9, True => 2 4 6 8 10␤»
GLRelia rakudo-moar 23b3d2: OUTPUT«True => 2 4 6 8 10False => 1 3 5 7 9␤»
colomon m: my %b := BagHash.new; say (1..10).classify(* %% 2, :into(%b)) 10:44
camelia rakudo-moar eab03c: OUTPUT«BagHash.new(False(5), True(5))␤»
GLRelia rakudo-moar 23b3d2: OUTPUT«BagHash.new(True(5), False(5))␤»
colomon lizmat: do those results look correct to you? :)
colomon did not change anything to get them.
timotimo jnthn: jimmyz suggested that and i've removed the macros, now i get an nfa with a num_state_edges that's set to an insane number d|
:|
lizmat colomon: looks sane to me? 10:45
colomon lizmat: groovy.
# expected: ("even"=>2,"odd"=>2).BagHash
# got: ("even"=>2,"odd"=>2).BagHash
timotimo oh
num_state_edges is actually supposed to be a pointer to an array 10:46
colomon is-deeply maybe doesn’t handle BagHashes?
10:46 Ven left
timotimo a pointer ending in a "1" doesn't seem like a sane pointer :| 10:46
lizmat does is-deeply handle different hash orders at all?
lunch& 10:47
timotimo is-deeply lunch 10:49
ShimmerFairy colomon: is-deeply does eqv, unless I'm mistaken, so maybe check that? 10:51
btyler_ I think I found something memory-leaky: gist.github.com/kanatohodets/b3692...a3e47620a1
FROGGS m: my @a = [$["foo", "bar"], "baz"]; say @a X=> 1 # how do I make that outputting "foo => 1bar => 1baz => 1" ? 10:52
camelia rakudo-moar a46b09: OUTPUT«Cannot convert string to number: base-10 number must begin with valid digits or '.' in '⏏foo' (indicated by ⏏)␤ in block <unit> at /tmp/heKhShyIC3:1␤␤Actually thrown at:␤ in block <unit> at /tmp/heKhShyIC3:1␤␤»
GLRelia rakudo-moar 23b3d2: OUTPUT«foo bar => 1baz => 1␤»
ShimmerFairy m: my @a = [$["foo", "bar"], "baz"]; say (@a X=> 1).perl
camelia rakudo-moar a46b09: OUTPUT«Cannot convert string to number: base-10 number must begin with valid digits or '.' in '⏏foo' (indicated by ⏏)␤ in block <unit> at /tmp/yQMaBQSapr:1␤␤Actually thrown at:␤ in block <unit> at /tmp/yQMaBQSapr:1␤␤»
GLRelia rakudo-moar 23b3d2: OUTPUT«($["foo", "bar"] => 1, :baz(1)).Seq␤»
jnthn timotimo: Sounds like memory corruption of some kind then... :S 10:53
|Tux| laben, /me is ready for some digging while a heavy process is running.
colomon timotimo: a pointer ending in 1 is just a bit odd, that’s all.
FROGGS hehe
even that, aye
btyler_ jnthn: FROGGS pointed me your way on that leaky thing -- anything I can do to narrow down the hunt?
[Coke] btyler_: verified, that seems to chew memory on my OS X box.
10:54 rindolf left, Sqirrel left
FROGGS jnthn: btw, he says it does not leak when $count is a constant... leaking code is: my $count = 0; loop { 1 if $count %% 1 } 10:54
so it can't be the %% op
nine colomon: BagHash or one of its base classes may need a WHICH 10:55
10:56 Sqirrel joined
jnthn Well, 0 %% 1 probably gets constant-folded 10:56
10:56 dayangkun left
jnthn m say 0 %% 1 10:56
|Tux| does «class C does Iterable does Positional {» in glr need additional roles to also do Hash lookups with AT-KEY ? 10:57
jnthn m: say 0 %% 1
camelia rakudo-moar a46b09: OUTPUT«True␤»
GLRelia rakudo-moar 23b3d2: OUTPUT«True␤»
jnthn |Tux|: Things you can hash-index into normally does Associative
timotimo i see what you did there, colomon :) 10:58
jnthn I'd probably valgrind it
|Tux| jnthn, and what should method "of" be implementing in that case? 10:59
colomon timotimo: I was disturbingly pleased by that one. ;)
jnthn |Tux|: What type constraint anything you store in it has; return Mu if you aren't implementing a typed data structure
11:00 RabidGravy left
laben |Tux|: i can send you a gist with the changes, too busy to investigate right now 11:00
|Tux| laben, yes please
11:00 rurban joined
laben |Tux|: here is it, they're all together gist.github.com/e3ab90806f82cd349493 11:04
|Tux| thanks
laben some seems precedence problems, maybe due to the slang
11:05 Ven joined
colomon m: %b := BagHash.new(); %b.classify-list( {.comb}, 20..40 ); say %b.perl 11:06
camelia rakudo-moar a46b09: OUTPUT«5===SORRY!5=== Error while compiling /tmp/5AlhUIubL2␤Variable '%b' is not declared␤at /tmp/5AlhUIubL2:1␤------> 3<BOL>7⏏5%b := BagHash.new(); %b.classify-list( {␤»
GLRelia rakudo-moar 23b3d2: OUTPUT«5===SORRY!5=== Error while compiling /tmp/z7RZQ2QFmg␤Variable '%b' is not declared␤at /tmp/z7RZQ2QFmg:1␤------> 3<BOL>7⏏5%b := BagHash.new(); %b.classify-list( {␤»
colomon m: my %b := BagHash.new(); %b.classify-list( {.comb}, 20..40 ); say %b.perl
camelia rakudo-moar a46b09: OUTPUT«postcircumfix:<{ }> not defined for type Int␤ in block <unit> at /tmp/jOo0Ksq1et:1␤␤Actually thrown at:␤ in any at src/gen/m-Metamodel.nqp:2867␤ in block <unit> at /tmp/jOo0Ksq1et:1␤␤»
GLRelia rakudo-moar 23b3d2: OUTPUT«Method 'succ' not found for invocant of class 'List'␤ in block <unit> at /tmp/TaCRMRZ234:1␤␤»
colomon lizmat: I think this operation is broken in both both nom and glr.
timotimo oh, duh 11:07
colomon lizmat: but then, I’m not 100% sure what it is supposed to do
timotimo if num_states is 0, of course the states array isn't defined memory
dalek kudo/glr: f761858 | jnthn++ | src/Perl6/Actions.nqp:
(1, 2 ; 3, 4) now makes a List of List.

Literally. No LoL.		 11:08
11:08 Ven left 11:09 colomon left
JimmyZ I thought LoL is short for List of List ... 11:10
tadzik we python now 11:11
jnthn JimmyZ: Now there's no type LoL though :) 11:12
sergot hi #perl6!
tadzik oh hai
11:13 llfourn joined, colomon joined
dalek rl6-roast-data: 74100de | coke++ | / (2 files):
another glr run for 'today'
 11:14
timotimo i forgot a single "re-grab the NFA pointer from the nfa root if the gc has kicked in since"
argh! 11:15
[Coke] m: 118901 - 118302 # of tests moar-jit is passing vs. moar-glr 11:16
camelia rakudo-moar a46b09: OUTPUT«WARNINGS:␤Useless use of "-" in expression "118901 - 118302" in sink context (line 1)␤»
GLRelia rakudo-moar 23b3d2: OUTPUT«WARNINGS:␤Useless use of "-" in expression "118901 - 118302" in sink context (line 1)␤»
[Coke] m: say 118901 - 118302 # of tests moar-jit is passing vs. moar-glr
camelia rakudo-moar a46b09: OUTPUT«599␤»
GLRelia rakudo-moar 23b3d2: OUTPUT«599␤»
timotimo the code to do that was actually totally bogus %)
[Coke] m: for (9...1) X (9...1) X (9...1) X (9...1) X (9...1) X (9,7,3) -> $a,$b,$c,$d,$e,$f {} 11:19
GLRelia rakudo-moar f76185: OUTPUT«Method 'pull-one' not found for invocant of class 'List'␤ in block <unit> at /tmp/P1lB7Eam_c:1␤␤»
camelia rakudo-moar a46b09: OUTPUT«Too few positionals passed; expected 6 arguments but got 3␤ in block <unit> at /tmp/z2z1NpgXqB:1␤␤»
dalek kudo/nom: 5d3fe6a | moritz++ | t/spectest.data:
Comment out IO-Socket-Async.t for the release
[Coke] ^^ jnthn - that's from your RT #114918
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=114918
colomon m: my %b := BagHash.new(); %b.classify-list( {.comb}, 20..40 ); say %b.perl 11:20
camelia rakudo-moar a46b09: OUTPUT«postcircumfix:<{ }> not defined for type Int␤ in block <unit> at /tmp/Oojcn28zVs:1␤␤Actually thrown at:␤ in any at src/gen/m-Metamodel.nqp:2867␤ in block <unit> at /tmp/Oojcn28zVs:1␤␤»
GLRelia rakudo-moar f76185: OUTPUT«Method 'succ' not found for invocant of class 'List'␤ in block <unit> at /tmp/oxvcsh2p4W:1␤␤»
colomon masak: ^^
masak submits rakudobug
colomon m: my %b; %b.classify-list( {.comb}, 20..40 ); say %b.perl 11:21
camelia rakudo-moar a46b09: OUTPUT«{"2" => {"0" => [20], "1" => [21], "2" => [22], "3" => [23], "4" => [24], "5" => [25], "6" => [26], "7" => [27], "8" => [28], "9" => [29]}, "3" => {"0" => [30], "1" => [31], "2" => [32], "3" => [33], "4" => [34], "5" => [35], "6" => [36], "7" => [37], "8" …»
GLRelia rakudo-moar f76185: OUTPUT«{"2" => ${"0" => $[20], "1" => $[21], "2" => $[22], "3" => $[23], "4" => $[24], "5" => $[25], "6" => $[26], "7" => $[27], "8" => $[28], "9" => $[29]}, "3" => ${"0" => $[30], "1" => $[31], "2" => $[32], "3" => $[33], "4" => $[34], "5" => $[35], "6" => $[36]…»
[Coke] r: []=5; say "alive" 11:22
GLRelia rakudo-moar f76185: OUTPUT«alive␤»
..rakudo-jvm f76185: OUTPUT«Can't open perl script "/home/steve/rakudo-inst/bin/eval-client.pl": No such file or directory␤»
camelia rakudo-{moar,jvm} a46b09: OUTPUT«alive␤»
11:22 Ven joined
masak m: my %b := BagHash.new(); say %b{'foo'} 11:24
camelia rakudo-moar a46b09: OUTPUT«0␤»
GLRelia rakudo-moar f76185: OUTPUT«0␤»
FROGGS m: say (42, 3).WHAT
camelia rakudo-moar a46b09: OUTPUT«(Parcel)␤»
GLRelia rakudo-moar f76185: OUTPUT«Invocant requires an instance of type List, but a type object was passed. Did you forget a .new?␤ in block <unit> at /tmp/7_bfpx7ZdU:1␤␤»
moritz huh. 11:25
[Coke] my @RT120506-bind := Array[Array[Bool]].new($(Array[Bool].new(True, False, True)), $(Array[Bool].new(True))); use Test; is-deeply @RT120506-bind[0, 1]».Parcel, ((True, False, True), (True,)), "urk?"
m: my @RT120506-bind := Array[Array[Bool]].new($(Array[Bool].new(True, False, True)), $(Array[Bool].new(True))); use Test; is-deeply @RT120506-bind[0, 1]».Parcel, ((True, False, True), (True,)), "urk?"
camelia rakudo-moar a46b09: OUTPUT«not ok 1 - urk?␤␤# Failed test 'urk?'␤# at /tmp/RUm0uDPc9K line 1␤# expected: $((Bool::True, Bool::False, Bool::True), (Bool::True,))␤# got: $(Array[Bool].new(Bool::True, Bool::False, Bool::True), Array[Bool].new(Bool::True))␤»
GLRelia rakudo-moar f76185: OUTPUT«Method 'Parcel' not found for invocant of class 'Bool'␤ in block <unit> at /tmp/DwdNCPC3u2:1␤␤»
laben m: for (1..2) X (1..2) -> $a, $b {}; say "OK"; for (1..9) X (1..9) -> $a, $b {}
camelia rakudo-moar a46b09: OUTPUT«OK␤»
GLRelia rakudo-moar f76185: OUTPUT«OK␤Too few positionals passed; expected 2 arguments but got 1␤ in block <unit> at /tmp/Lw8XlPGktm:1␤␤»
laben why ^^^ is it different between 1..2 and 1..9? 11:26
11:26 CIAvash left
colomon what does 1..3 do? 11:26
masak laben: even vs odd number of elements
laben: you're looping over the tuples themselves, and there are 4 and 81 11:27
laben masak: why does it even matter, shouldn't there be a single list created by X?
[Coke] m: my @m <== sort <== ("c", "b", "a") ; say @m;
camelia rakudo-moar a46b09: OUTPUT«5===SORRY!5=== Error while compiling /tmp/RbxmL4VmsT␤Unable to parse expression in quote words; couldn't find final '>'␤ ␤at /tmp/RbxmL4VmsT:1␤------> 3m <== sort <== ("c", "b", "a") ; say @m;7⏏5<EOL>␤ expecting any of:␤ …»
GLRelia rakudo-moar f76185: OUTPUT«5===SORRY!5=== Error while compiling /tmp/82aR3aEFpM␤Unable to parse expression in quote words; couldn't find final '>'␤ ␤at /tmp/82aR3aEFpM:1␤------> 3m <== sort <== ("c", "b", "a") ; say @m;7⏏5<EOL>␤ expecting any of:␤ …»
masak nope, not without `flat`
laben doesnt X create a list of tuples?
masak that's what's meant by "learn to love `flat`" :)
laben: yes, it does. that's what you're seeing.
[Coke] m: my @m <== sort() <== ("c", "b", "a") ; say @m; 11:28
camelia rakudo-moar a46b09: OUTPUT«a b c␤»
GLRelia rakudo-moar f76185: OUTPUT«abc␤»
masak laben: $a and $b each get such a *tuple*
[Coke] ^^ is it reasonable to require the parens there? RT #121843
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=121843
masak m: for (1..2) X (1..2) -> $a, $b { say $a }
camelia rakudo-moar a46b09: OUTPUT«1␤1␤2␤2␤»
GLRelia rakudo-moar f76185: OUTPUT«1 1␤2 1␤»
masak oh, it differs between cam and GLR
laben m: for $((1..2) X (1..2)) -> $a { dd $a };
camelia rakudo-moar a46b09: OUTPUT«$a = $(1, 1, 1, 2, 2, 1, 2, 2)␤»
GLRelia rakudo-moar f76185: OUTPUT«$a = $(1, 1)␤$a = $(1, 2)␤$a = $(2, 1)␤$a = $(2, 2)␤»
masak [Coke]: reasonable to require them? I dunno. they don't look required to me. 11:29
[Coke]: <== has very low precedence
certainly lower than comme
comma*
dalek ast/glr: 5d29822 | colomon++ | S32-list/classify.t:
Add some very minimal tests for Baggy classify.

As far as I could tell, there weren't any tests for this in roast so far.
[Coke] masak: required to get the "desired" output. If the original request is borked, please reject the ticket. 11:30
masak looks at ticket 11:31
dalek pan style="color: #395be5">perl6-examples: 0c7e431 | paultcochrane++ | / (3 files):
Add HTTP::Easy::PSGI as a dependency
[Coke] masak++
timotimo oh yay! 11:32
laben masak: thanks! your example helped me understand
timotimo edges of 18: 2 Internal error: invalid thread ID -334888944 in GC work pass
masak [Coke]: oh, the parens on `sort()` ? 11:33
[Coke] masak: Hai.
masak laben: running code beats most things in terms of explanatory power :)
[Coke]: I... dunno. I usually put those parens in, "because then it works"
[Coke]: I don't actually know if it should work without them. kinda feels like it should. 11:34
[Coke]: simply because `sort` === `sort()`
[Coke] Ditto
nine m: for (1..2) X (1..2) -> ($a, $b) { say $a } # explicit deconstruction
camelia rakudo-moar a46b09: OUTPUT«Too few positionals passed; expected 2 arguments but got 0 in sub-signature␤ in block <unit> at /tmp/EM1I8Qz7Ih:1␤␤»
GLRelia rakudo-moar f76185: OUTPUT«1␤1␤2␤2␤»
masak nine++ 11:35
m: say 42.indent(4) 11:37
camelia rakudo-moar a46b09: OUTPUT«Method 'indent' not found for invocant of class 'Int'␤ in block <unit> at /tmp/kCES7V68L9:1␤␤»
GLRelia rakudo-moar f76185: OUTPUT«Method 'indent' not found for invocant of class 'Int'␤ in block <unit> at /tmp/zmnIOfxTcT:1␤␤»
dalek kudo/nom: 47ddca1 | moritz++ | docs/ChangeLog:
Add some more changelog entries
masak should .indent be on Cool?
timotimo m: say -1073741810.base(16)
camelia rakudo-moar a46b09: OUTPUT«Cannot convert string to number: trailing characters after number in '3⏏FFFFFF2' (indicated by ⏏)␤ in block <unit> at /tmp/ny19SH5eIL:1␤␤Actually thrown at:␤ in block <unit> at /tmp/ny19SH5eIL:1␤␤»
GLRelia rakudo-moar f76185: OUTPUT«Cannot convert string to number: trailing characters after number in '3⏏FFFFFF2' (indicated by ⏏)␤ in block <unit> at /tmp/y0jUPFzR_i:1␤␤Actually thrown at:␤ in block <unit> at /tmp/y0jUPFzR_i:1␤␤»
[Coke] wouldn't mind if it was string only. anything not cool is just going to be a single line, no?
moritz masak: that would be consistent with how we do it usually
timotimo m: say "-1073741810".base(16)
camelia rakudo-moar a46b09: OUTPUT«Method 'base' not found for invocant of class 'Str'␤ in block <unit> at /tmp/CoNGQxN2Fc:1␤␤»
GLRelia rakudo-moar f76185: OUTPUT«Method 'base' not found for invocant of class 'Str'␤ in block <unit> at /tmp/lqs6JXuy1N:1␤␤»
dalek kudo/glr: f09891f | colomon++ | src/core/Baggy.pm:
Hash.classify-list fix to Baggy.classify-list.

Unfortunately, even in nom this part of the classify-list functionality didn't work. This fix makes it fail like nom does, rather than the entire new way of failing that glr had discovered.
timotimo ... huh?
timotimo m: say (-1073741810).base(16) 11:38
camelia rakudo-moar a46b09: OUTPUT«-3FFFFFF2␤»
GLRelia rakudo-moar f76185: OUTPUT«-3FFFFFF2␤»
masak ok, if someone wants an LHF:
spectest and move Str.indent to Cool.indent :)
colomon m: say "-1073741810”.Int.base(16)
camelia rakudo-moar a46b09: OUTPUT«5===SORRY!5=== Error while compiling /tmp/1OCZvmdXWF␤Unable to parse expression in double quotes; couldn't find final '"' ␤at /tmp/1OCZvmdXWF:1␤------> 3say "-1073741810”.Int.base(16)7⏏5<EOL>␤ expecting any of:␤ argument lis…»
GLRelia rakudo-moar f76185: OUTPUT«5===SORRY!5=== Error while compiling /tmp/wmGNaJlTAN␤Unable to parse expression in double quotes; couldn't find final '"' ␤at /tmp/wmGNaJlTAN:1␤------> 3say "-1073741810”.Int.base(16)7⏏5<EOL>␤ expecting any of:␤ argument lis…»
dalek pan style="color: #395be5">perl6-examples: 0714a62 | paultcochrane++ | categories/cookbook/19cgi-programming/19-01cgi-script.pl:
Purge trailing whitespace
pan style="color: #395be5">perl6-examples: f87f75e | paultcochrane++ | categories/cookbook/ (2 files):
Put description text after AUTHOR

Now the parser can find the example author information.
colomon m: say “-1073741810”.Int.base(16)
camelia rakudo-moar a46b09: OUTPUT«-3FFFFFF2␤»
GLRelia rakudo-moar f76185: OUTPUT«-3FFFFFF2␤»
ab6tract is excited to leave for the airport to a couple hours :) 11:44
*in 11:45
colomon m: sub subber ($a) { $a - ($a % 10) }; my @list = 29, 7, 12, 9, 18, 23, 3, 7; say categorize(&subber, @list) 11:46
camelia rakudo-moar a46b09: OUTPUT«0 => 7 9 3 7, 10 => 12 18, 20 => 29 23␤»
GLRelia rakudo-moar f09891: OUTPUT«Cannot call categorize-list(Hash[Any,Any]: Sub, Array); none of these signatures match:␤ (Hash $: %test, $list, *%_)␤ (Hash $: @test, $list, *%_)␤ in block <unit> at /tmp/yn6E8RbDFm:1␤␤»
Ven ab6tract: are you going to yapc::eu? :)
11:47 Ven left
FROGGS nine: besides that I think the Array->List conversion makes sense for Signature.named_names, I just need to .Slip in that map in MAIN_HELPER 11:47
nine FROGGS: sounds good :) 11:48
11:48 virtualsue joined 11:52 brrt joined, rurban left
[ptc] itz: you've got a vim process on hack going nuts. 11:53
itz: looks like it's hanging; it's using ~100% cpu and has been for a while
itz: oddly enough, I had such a process just now as well. Had to kill it and then all was good
ab6tract Ven: SPW today, YAPC::EU on Tuesday :D
tadzik is yapc::eu so soon? 11:54
indeed :)
muraiki is doc.perl6.org/language/containers still up to date? I thought I had seen talk about changing/eliminating parcels or something a while back 11:55
ab6tract ('the hackathon' xx *)>>.++
moritz muraiki: it's up to date with respect to current rakudo/nom; GLR changes things. 11:56
muraiki moritz: thanks
ab6tract muraiki: indeed. the very last details seem to be getting hammered out as we speak
muraiki woot :)
itz [ptc]: I've just killed all those process I was running a cron which was an old snapshot of the perl6-examples update processes 11:57
moritz is LoL officially dead? 11:58
itz [ptc]: I think I will put a timeout in the code .. perhaps perl6 syntax is sometimes too much for vim syntax hilighting
dalek kudo/glr: 285febd | jnthn++ | src/ (4 files):
Stub in new [; ] and {; } handling.

All mutli-dispatch errors for now.
moritz itz: fwiw I've added a CPU time limit in the updated script, but it applies globally
11:59 Ven joined
Ven absolutely <3 OO::* 11:59
nine moritz: yes
dalek kudo/glr: a70367c | jnthn++ | src/core/A (2 files):
Start preparing for multi-dim *-POS.

We'll pass multiple indices to the [AT|ASSIGN|EXISTS...]-POS methods. This will allow things that know how many dimensions they have to just bind them directly to parameters, which will be efficient (and allow the dynamic optimizer to inline, etc.). We'll be able to do some case analysis on the indexer itself at optimization time to eliminte the LoL in the commonest cases also.
This patch simply adds fallbacks for the multi-indice case that dig into the data structure a level at a time.		 12:04
kudo/glr: 6b75f33 | jnthn++ | src/core/Any.pm:
Fix a missing "is rw".
FROGGS nine: there are more code paths that expect an Array but not a List sadly :o( 12:06
nine: I'll care about this later I think
dalek ast/glr: 3cdd87a | moritz++ | S04-statements/gather.t:
Remove .lol from gather.t
rl6-roast-data: 3cfb0b3 | coke++ | / (2 files):
another glr run for 'today'
 12:08
12:08 itz is now known as _itz 12:09 itz joined
|Tux| if I have a to_list that is implemented as a gather, .flat on that list in glr will take forever 12:10
as in it is not lazy at all
laben, that is the basic cause of failure for t/78
and adding .flat is not helping there 12:11
dalek ast: 4584bfe | ShimmerFairy++ | S02-literals/allomorphic.t:
Add tests for allomorphic types (IntStr etc.)

Unfortunately they all have to be fudged for rakudo for the time being.
12:11 domidumont left
colomon in GLR, are there gotchas about using push? 12:12
lizmat colomon: yes
colomon: and we're thinking about them 12:13
colomon is ready to recieve wisdom
…. oh
lizmat but no solution just yet :-(
or resolution, I should say
12:15 domidumont joined
moritz why not just git push a **@elems signature, and consider everything on the first level a new element to add? 12:15
m: sub f(**@e) { @e.perl }; say f 1..3
camelia rakudo-moar 47ddca: OUTPUT«(1, 2, 3)␤»
GLRelia rakudo-moar 6b75f3: OUTPUT«[1..3]␤»
moritz m: sub f(**@e) { @e[0].perl }; say f 1..3 12:16
camelia rakudo-moar 47ddca: OUTPUT«1..3␤»
GLRelia rakudo-moar 6b75f3: OUTPUT«1..3␤»
[Coke] we have a pending PR for nqp that's a few months old: github.com/perl6/nqp/pull/228
moritz though there are probably implications that I've missed (re .push) 12:17
TimToady that is one of the proposed solutions, but when I tried it, the setting blew up :)
12:17 rarara joined
moritz TimToady: that's decidedly LTA :-) 12:17
dalek kudo/glr: 496300b | lizmat++ | src/core/Backtrace.pm:
Fix RAKUDO_VERBOSE_STACKFRAME option
 12:20
12:21 araujo_ joined
timotimo as sort of expected, the MVM_nfa_to_statelist seems to be correct semantically, but it's not gc-clean :| 12:23
jnthn timotimo: Give it a GC bath 12:24
colomon ’s problem didn’t have anything to do with push, it turned out. :)
dalek kudo/glr: f2b993c | jnthn++ | src/core/array_slice.pm:
Make basic ; array indexing work again.

Just for simple indexes, not slices.
12:24 araujo left
jdv79 so, i have a fix for cpan but i'm not sure if the p5/p6 ecosystem sharing thing is well thought out. is there a consolidated info somewhere about this? 12:25
FROGGS jdv79: probably only in my brain
jdv79 can that be accessed? 12:26
FROGGS indirectly
there is a brain to text api you can use
timotimo yeah, you grab froggs at the back of his head and gently push it against the keyboard 12:27
lizmat so those are not freckles, but key impressions ?
FROGGS really gently please :o) 12:28
*g*
good that the L key is not at the middle top
dalek p: 5146561 | coke++ | t/docs/tests.t:
Don't use substr2 where substr will do.

Close #223		 12:30
jdv79 so it appears that most p5 ecosystemm things treat the dist/module/package names as unique names so what was the thinking on how that would be addreseed? 12:36
tadzik we don't :)
FROGGS lizmat: I can't spot in the clogs that you states that your patch causes regressions
tadzik panda recently stopped doing that 12:37
ie we have 2 modules named Foo in the ecosystem
jdv79 well, if we want to take advantage of the p5 stuff that's a problem
lizmat FROGGS: but I never stated that... because I didn't know it caused a regression
ShimmerFairy We generally don't, we're just using a P5-coded system :) 12:38
FROGGS lizmat: can you please check that? I loose again almost an hour :o(
lizmat in fact, I thought it fixed several tests and we were *pro*gressing
:-(
FROGGS s/check that/chek that next time/
nine Are we sure it's really regressions and not just uncovered bugs?
FROGGS nine: looks like regression to me 12:39
nine FROGGS: I'm curious. Is there a golf for reproducing the new error? 12:40
FROGGS m: my @a = ("array", "of", "params"); note @a 12:41
camelia rakudo-moar 47ddca: OUTPUT«array of params␤»
GLRelia rakudo-moar f2b993: OUTPUT«arrayofparams␤»
lizmat m: my @a = ("array", "of", "params"); say @a
camelia rakudo-moar 47ddca: OUTPUT«array of params␤»
GLRelia rakudo-moar f2b993: OUTPUT«arrayofparams␤»
dalek pan style="color: #395be5">perl6-examples: b1f0b92 | (Steve Mynott)++ | / (2 files):
HTTP::Easy::PSGI in ecosystem HTTP::Easy
FROGGS it might be just that the remaining note() candidate needs to be note(**@) instead of note(*@)
lizmat m: my @a = ("array", "of", "params"); say @a, 12:43
camelia rakudo-moar 47ddca: OUTPUT«array of params␤»
GLRelia rakudo-moar f2b993: OUTPUT«arrayofparams␤»
lizmat m: my @a = ("array", "of", "params"); say (@a,)
camelia rakudo-moar 47ddca: OUTPUT«array of params␤»
GLRelia rakudo-moar f2b993: OUTPUT«arrayofparams␤»
nine m: my @ac ("array", "of", "params"); say "@a";
camelia rakudo-moar 47ddca: OUTPUT«5===SORRY!5=== Error while compiling /tmp/K1pOSCRa8v␤Signatures as constraints on variables not yet implemented. Sorry. ␤at /tmp/K1pOSCRa8v:1␤------> 3my @ac ("array", "of", "params")7⏏5; say "@a";␤»
GLRelia rakudo-moar f2b993: OUTPUT«5===SORRY!5=== Error while compiling /tmp/ddLirgsRQK␤Signatures as constraints on variables not yet implemented. Sorry. ␤at /tmp/ddLirgsRQK:1␤------> 3my @ac ("array", "of", "params")7⏏5; say "@a";␤»
nine m: my @a = ("array", "of", "params"); say "@a";
camelia rakudo-moar 47ddca: OUTPUT«@a␤»
GLRelia rakudo-moar f2b993: OUTPUT«@a␤»
nine m: my @a = ("array", "of", "params"); say "{@a}";
camelia rakudo-moar 47ddca: OUTPUT«array of params␤»
GLRelia rakudo-moar f2b993: OUTPUT«array of params␤»
nine m: my @a = ("array", "of", "params"); say @a;
camelia rakudo-moar 47ddca: OUTPUT«array of params␤»
GLRelia rakudo-moar f2b993: OUTPUT«arrayofparams␤»
ShimmerFairy m: my @a = ("array", "of", "params"); say "@a[]";
camelia rakudo-moar 47ddca: OUTPUT«array of params␤»
GLRelia rakudo-moar f2b993: OUTPUT«array of params␤»
nine m: my @a = ("array", "of", "params"); say "array", "of", "params";
camelia rakudo-moar 47ddca: OUTPUT«arrayofparams␤»
GLRelia rakudo-moar f2b993: OUTPUT«arrayofparams␤»
nine m: my @a = ("array", "of", "params"); say $@a; 12:44
camelia rakudo-moar 47ddca: OUTPUT«array of params␤»
GLRelia rakudo-moar f2b993: OUTPUT«array of params␤»
jdv79 i only see one foo on modules.perl6.org
FROGGS jdv79: we need to fix the ecosystem code and panda to support multiple same named dists 12:45
nine m: my @a = ("array", "of", "params"); say @a.gist;
camelia rakudo-moar 47ddca: OUTPUT«array of params␤»
GLRelia rakudo-moar f2b993: OUTPUT«array of params␤»
lizmat m: my @a = ("array", "of", "params"); say @a.gist
camelia rakudo-moar 47ddca: OUTPUT«array of params␤»
GLRelia rakudo-moar f2b993: OUTPUT«array of params␤»
tadzik I just noticed that modules.perl6.org doesn't support it
FROGGS I was always under the impression that 'say @a' implies 'say @a.gist' 12:46
and I vote for that 'say 1, 2' always has been wrong 12:47
nine say @a meaning say @a.gist was probably one of the many rules that we replaced by the single-arg-rule
FROGGS true, though I want to have a space between the elems
nine m: say 1, 2, 3; 12:49
camelia rakudo-moar 47ddca: OUTPUT«123␤»
GLRelia rakudo-moar f2b993: OUTPUT«123␤»
12:49 RabidGravy joined
FROGGS TimToady: can we please have a space character in there? 12:49
dalek kudo/glr: 2ccea71 | lizmat++ | src/core/io_operators.pm:
Revert "Remove say(\x) / note(\x) candidates"
osystem: 7f151ba | (Nova Patch)++ | META.list:
CLDR::List moved on GitHub
 12:50
[Coke] fixed the nqp docs tests so they are runnable again. Lots of docs missing +/or untested opcodes.
ShimmerFairy FROGGS: however, say "Foo: ", something-to-print($arg) will end up breaking (by putting extra spaces where you don't mean them) 12:51
[Coke] colomon opened a new RT? aiiigh. :) 12:52
|Tux| GLR: method to_list { gather { take 3 .. Inf }} ← that now returns a Range instead of a lazy list. Is that intended?
FROGGS m: say 12, 3; say (my @a = 12, 3); # I want bot output "12 3" 12:53
camelia rakudo-moar 47ddca: OUTPUT«123␤12 3␤»
GLRelia rakudo-moar f2b993: OUTPUT«123␤123␤»
FROGGS TimToady / lizmat: ^^
both* 12:54
[Coke] m: say 12, 3; say (my @a = 12, 3).perl
camelia rakudo-moar 47ddca: OUTPUT«123␤[12, 3]<>␤»
GLRelia rakudo-moar f2b993: OUTPUT«123␤[12, 3]␤»
[Coke] do you want default join() changed or default gist?
lizmat nothing to do with join 12:55
FROGGS [Coke]: neither
say calls .gist on the elems of the input array, and joins it with ''
ShimmerFairy like I said, it would break a lot of literal arg lists passed to 'say' that expect nothing to be placed between elements, so I don't think changing that is the right move.
FROGGS and this is wrong IMO
ShimmerFairy: I'm thinking about it 12:56
ShimmerFairy I think the better solution would be something that preserved the fact that you're passing a list object to say (isn't **@a supposed to do that?) 12:57
|Tux| m: class C{method foo{gather { take 3..Inf}}};C.new.foo[^9].perl.say 12:58
camelia rakudo-moar 47ddca: OUTPUT«(3, 4, 5, 6, 7, 8, 9, 10, 11)␤»
GLRelia rakudo-moar f2b993: OUTPUT«(3..Inf)␤»
[Coke] r: my @a=<a a a>;my @b=<b b b>;for (@a »,« @b) »,« @a -> $a, $b,$c {say "$a $b $c"}
GLRelia rakudo-jvm 2ccea7: OUTPUT«Can't open perl script "/home/steve/rakudo-inst/bin/eval-client.pl": No such file or directory␤»
..rakudo-moar 2ccea7: OUTPUT«Lists on either side of non-dwimmy hyperop of infix:<,> are not of the same length␤left: 2 elements, right: 1 elements␤ in block <unit> at /tmp/tmpfile:1␤␤»
timotimo m: say rand 5
GLRelia rakudo-moar 2ccea7: OUTPUT«5===SORRY!5=== Error while compiling /tmp/ploXygyYbC␤Unsupported use of rand(N); in Perl 6 please use N.rand for Num or (^N).pick for Int result␤at /tmp/ploXygyYbC:1␤------> 3say rand7⏏5 5␤»
camelia rakudo-{moar,jvm} 47ddca: OUTPUT«Lists on either side of non-dwimmy hyperop of infix:<,> are not of the same length␤left: 6 elements, right: 3 elements␤ in block <unit> at /tmp/tmpfile:1␤␤»
rakudo-moar 47ddca: OUTPUT«5===SORRY!5=== Error while compiling /tmp/od5piU737C␤Unsupported use of rand(N); in Perl 6 please use N.rand for Num or (^N).pick for Int result␤at /tmp/od5piU737C:1␤------> 3say rand7⏏5 5␤»
|Tux| is that glr behavior expected or shall nom and glr be the same? 12:59
[Coke] r: my @a := 1..*; @a := @a[1, 3 ... *].list 13:00
GLRelia rakudo-moar 2ccea7: OUTPUT«(timeout)»
..rakudo-jvm 2ccea7: OUTPUT«Can't open perl script "/home/steve/rakudo-inst/bin/eval-client.pl": No such file or directory␤»
camelia rakudo-jvm 47ddca: OUTPUT«␤(timeout)␤»
..rakudo-moar 47ddca: OUTPUT«(timeout)»
masak <FROGGS> TimToady: can we please have a space character in there? 13:01
FROGGS: I'm pretty sure I wouldn't want that
FROGGS masak: also not for 'say @a' ?
masak FROGGS: I use `say thing, thing, thing;` when I want to print things next to each other, no space.
dalek kudo/glr: 99f158a | TimToady++ | src/core/Array.pm:
omit redundant $ inside Array.perl
masak FROGGS: I see `say @a` as a different case.
FROGGS masak: okay, I can live with that
masak: so we want the old behaviour? 13:02
masak FROGGS: if I was the Emperor of Perl 6, `say @a` would actually print something like "[1, 2, 3]". but TimToady has already overruled that.
FROGGS: and so I'll have to contend with "1 2 3"
FROGGS k
dalek p/nfa_to_statelist: 2e95469 | timotimo++ | src/ (13 files):
test nfatostatelist

by calling it on every nfa we call nfafromstatelist on.
moritz m: my @a = 1 2, 3; say @a
camelia rakudo-moar 47ddca: OUTPUT«5===SORRY!5=== Error while compiling /tmp/4pb6s5CeRH␤Two terms in a row␤at /tmp/4pb6s5CeRH:1␤------> 3my @a = 17⏏5 2, 3; say @a␤ expecting any of:␤ infix␤ infix stopper␤ postfix␤ statement end␤ …»
GLRelia rakudo-moar 2ccea7: OUTPUT«5===SORRY!5=== Error while compiling /tmp/8Dnazf07Ct␤Two terms in a row␤at /tmp/8Dnazf07Ct:1␤------> 3my @a = 17⏏5 2, 3; say @a␤ expecting any of:␤ infix␤ infix stopper␤ postfix␤ statement end␤ …»
moritz m: my @a = 1, 2, 3; say @a
camelia rakudo-moar 47ddca: OUTPUT«1 2 3␤»
GLRelia rakudo-moar 2ccea7: OUTPUT«1 2 3␤»
FROGGS lizmat / TimToady: so given ShimmerFairy's and masak's opinions, I want to old behaviour back 13:03
moritz I expected that to say 123, because say() has a slurpy arg, no?
masak you mean you expected that under GLR?
dalek kudo/glr: f26f52d | (Stefan Seifert)++ | src/core/Supply.pm:
Fix Supply.rotor on non-lazy lists

Fixes "Type check failed in binding; expected 'Positional' but got
  'Seq'" in rotor caused by xx returning a Seq now and we try to bind to
an @array. Need to .flat that because the code expects a flattened list of values instead of a list of lists.
p: 34501ae | coke++ | t/docs/tests.t:
make t/docs/tests.t runnable again
FROGGS lizmat / TimToady: so, space for passing an @a, no space for literal lists
moritz: don't mind the implementation
13:03 rurban joined
[Coke] r: role c {}; my @a = 1,2,3,4; my $b = @a.pick(2) but c; say $b.WHAT; say $b; 13:03
GLRelia rakudo-moar 2ccea7: OUTPUT«(Seq+{c})␤2 4␤»
..rakudo-jvm 2ccea7: OUTPUT«Can't open perl script "/home/steve/rakudo-inst/bin/eval-client.pl": No such file or directory␤»
camelia rakudo-moar 47ddca: OUTPUT«(List+{c})␤This type does not support elems␤ in block <unit> at /tmp/tmpfile:1␤␤» 13:04
..rakudo-jvm 47ddca: OUTPUT«(List+{c})␤java.lang.RuntimeException: This type does not support positional operations␤ in block <unit> at /tmp/tmpfile:1␤␤»
ShimmerFairy masak: there can be brackets at least, so we know how things are nested :)
laben |Tux|: i'm still at work for some hours, but did you try s/take $thing/take flat $thing/ ?
FROGGS gah, everything boils down to the fact that I like and dislike the one-arg-rule
masak ShimmerFairy: aye, that would be a step up 13:05
|Tux| laben. .flat will cause a hang on .inf
masak ShimmerFairy: "[1 2 3]" then?
jdv79 mulitple opinion disorder?
masak jdv79: multiple opinions -- the first step to finding a good solution. 13:06
laben |Tux|: then "take eager $thing" should avoid it
or take $thing.eager
nine I really don't see why say $@array; is not enough. You will get the spaces and my @array = 1, 2, 3; say @array does exactly the same as say 1, 2, 3;
FROGGS nine: I'm not saying I cannot be convinced, I just say I don't like it atm... 13:07
ShimmerFairy masak: yeah, particularly for things like (1 2 (3 4 5)) they'd be nice. Brackets are probably debatable (since it's .gist and not .perl), and I'm half-tempted to suggest some pair of brackets that aren't used for anything in standard P6 code :P (sorta like how we use corner quotes in Matches, except they _are_ valid P6) 13:08
TimToady I think I've already said I'm fine with brackets for .gist
at least for sub-gists 13:09
13:09 brrt left
FROGGS .oO( say quickly: sub-gists sub-gists sub-gists ) 13:09
ShimmerFairy TimToady: yeah, that's what I was pointing out, you being fine with .gist brackets :) 13:10
13:12 skids joined 13:13 telex left 13:14 rmgk_ joined, rmgk is now known as Guest88326, Guest88326 left, rmgk_ is now known as rmgk
dalek pan style="color: #395be5">perl6-examples: fb858cb | (Steve Mynott)++ | .travis.yml:
fix travis dep
 13:14
13:14 telex joined
Ven masak: github.com/vendethiel/6meta-experiments is my church.pl :) 13:17
btyler_ jnthn: bingo! that was it -- shifting clear_temp_bigints out of the divzero conditional stopped the leak
tadzik if a sub has a named parameter as a default value, is there any value I can pass so that it chooses the default over it? 13:18
I'm trying to avoid having two calls in my code, one with the named argument and the other without
timotimo you can |%(...) to get that
colomon [Coke]: found code in rakudo that never worked. I’m not even sure anyone knows what it should do… 13:19
timotimo those things are there sometimes 13:20
masak m: sub foo(:$bar = 42) { say $bar }; foo(bar => Nil) 13:22
camelia rakudo-moar 47ddca: OUTPUT«Nil␤»
GLRelia rakudo-moar f26f52: OUTPUT«Nil␤»
masak m: sub foo(:$bar = 42) { say $bar }; foo(bar => Empty)
camelia rakudo-moar 47ddca: OUTPUT«Empty␤»
GLRelia rakudo-moar f26f52: OUTPUT«␤»
masak m: sub foo(:$bar = 42) { say $bar }; foo(bar => Int)
camelia rakudo-moar 47ddca: OUTPUT«(Int)␤»
GLRelia rakudo-moar f26f52: OUTPUT«(Int)␤»
masak tadzik: doesn't look like it
m: sub foo(:$bar = 42) { say $bar }; my %h; foo(|%h) 13:23
camelia rakudo-moar 47ddca: OUTPUT«42␤»
GLRelia rakudo-moar f26f52: OUTPUT«42␤»
masak m: sub foo(:$bar = 42) { say $bar }; my %h = False ?? bar => 42 !! (); foo(|%h)
camelia rakudo-moar 47ddca: OUTPUT«42␤»
GLRelia rakudo-moar f26f52: OUTPUT«42␤»
dalek kudo/glr: 4814073 | (Stefan Seifert)++ | src/core/io_operators.pm:
Revert "Revert "Remove say(\x) / note(\x) candidates""

This reverts commit 2ccea71e813c98e42c27d0da77f0a46878c197d8.
The original change actually is correct and consistent with the single-arg-rule of GLR.
kudo/glr: 041bf0f | (Stefan Seifert)++ | src/core/io_operators.pm:
Avoid flattening of say's arguments

This makes say @a, @b; mean say @a.gist, @b.gist; which would probably be what users expect.
masak m: sub foo(:$bar = 42) { say $bar }; my %h = True ?? bar => 5 !! (); foo(|%h)
camelia rakudo-moar 47ddca: OUTPUT«5␤»
GLRelia rakudo-moar f26f52: OUTPUT«5␤»
masak tadzik: that seems to work. timotimo++
dalek kudo/glr: 965539c | FROGGS++ | src/core/Main.pm:
handle param aliases in MAIN_HELPER

This fixes the case where an alias to a named param existed, and the usage message was printed as a false positive when passing either the
  param or the alias.		 13:25
FROGGS okay, now I was able to check that my path does not do any harm 13:27
nine :)
FROGGS :o)
13:27 llfourn left
FROGGS nine: that means we will adjust note.t and say.t? 13:28
nine FROGGS: I'll have a look at them, yes. 13:29
FROGGS nine: I currently do 13:30
colomon when you guys have been counting the number of GLR test failures, was that counting categorize.t (which crashed completely, rather than printing not oks)?
timotimo masak: you know what confused me for a moment there? :)
dalek ast/glr: cb71834 | FROGGS++ | S32-io/note.t:
adjust expectation of "note @a"
 13:32
pan style="color: #395be5">perl6-examples: 29ea342 | (Steve Mynott)++ | / (4 files):
remove one list of dependencies and doc the others
 13:34
FROGGS m: say "o", "k 2 - say with multiple parame", "ters (1)"
camelia rakudo-moar 47ddca: OUTPUT«ok 2 - say with multiple parameters (1)␤»
GLRelia rakudo-moar 965539: OUTPUT«ok 2 - say with multiple parameters (1)␤»
nine FROGGS: please hold your horses!
ugexe RecommendationManager is such a long name :(
nine My **@args change made it regress to before lizmat++'s change
FROGGS ohh! 13:35
jdv79 RecoMan?
FROGGS boy! you! :D
13:35 ab6tract left
colomon m: my @list = 29, 7, 12, 9, 18, 23, 3, 7; say categorize( [ 0 xx 10, 10 xx 10, 20 xx 10 ], @list).perl 13:35
camelia rakudo-moar 47ddca: OUTPUT«Hash[Any,Any].new(0 => [7, 9, 3, 7], 10 => [12, 18], 20 => [29, 23])␤»
GLRelia rakudo-moar 965539: OUTPUT«Cannot call categorize-list(Hash[Any,Any]: Block, Array); none of these signatures match:␤ (Hash $: %test, $list, *%_)␤ (Hash $: @test, $list, *%_)␤ in block <unit> at /tmp/LfBz7gVKSN:1␤␤»
tadzik ugexe: yt? 13:38
ugexe yes
tadzik awesome
want to test a fix for github.com/tadzik/panda/issues/142?
ugexe: it's at github.com/tadzik/panda/tree/gh-142 13:39
13:40 breinbaas left
FROGGS m: say 42 for Nil # everybody agree that this should iterate onec? 13:42
camelia rakudo-moar 47ddca: OUTPUT«42␤»
GLRelia ( no output )
ugexe i will, but im not sure where i'll have something to smoke with
FROGGS once, even
ugexe when rather
tadzik you mean a module, or hardware?
13:43 itz left
ugexe hardware. which i have, its just working atm heh 13:43
tadzik oh, okay :)
FROGGS m: say 42 for Int; say Int.elems; # ohh! 13:44
camelia rakudo-moar 47ddca: OUTPUT«42␤0␤»
GLRelia rakudo-moar 965539: OUTPUT«0␤»
tadzik FROGGS: reactionimage.org/img/gallery/1968804931.jpg
FROGGS jnthn: can you tell me that this is correct or not?
tadzik s/65/42/
FROGGS tadzik++ 13:45
masak m: say 42 for Int
camelia rakudo-moar 47ddca: OUTPUT«42␤»
GLRelia ( no output )
masak m: Int.map({ say 42 })
camelia rakudo-moar 47ddca: OUTPUT«42␤»
GLRelia ( no output )
masak Int-eresting...
FROGGS masak: its .elems is zero, so there is no iteration
masak yes, but whyyyyyyyyyyyyy 13:46
FROGGS though I wonder... yes, exactly
masak m: say Int.elems
camelia rakudo-moar 47ddca: OUTPUT«0␤»
GLRelia rakudo-moar 965539: OUTPUT«0␤»
TimToady m: say Any.elems
camelia rakudo-moar 47ddca: OUTPUT«0␤»
GLRelia rakudo-moar 965539: OUTPUT«0␤»
masak maybe the only way to get an answer to that is to spectest a change to it
laben m: say 42 for Empty 13:48
camelia ( no output )
GLRelia ( no output )
laben m: say 42 for ()
camelia ( no output )
GLRelia ( no output )
jnthn I'm pretty sure Int.elems should say 1
laben m: say 42 for (Any,)
camelia rakudo-moar 47ddca: OUTPUT«42␤»
GLRelia rakudo-moar 965539: OUTPUT«42␤»
jnthn Because it should be equivalent to Int.list.elems
13:48 llfourn joined
jnthn And Int.list should be (Int,) 13:48
masak ooh, maybe a blame can answer it?
timotimo moritz: + with/orwith/without conditionals allow you to check for .defined but
FROGGS /home/froggs/dev/rakudo/src/core/Any.pm:93: multi method elems(Any:U:) { 0 } 13:49
/home/froggs/dev/rakudo/src/core/Any.pm:94: multi method elems(Any:D:) { self.list.elems }
timotimo it looks like the supply/whenever line slipped in between
masak yep 13:50
6a3fc316b185057eee161d210493e8a442ec0a22
github.com/rakudo/rakudo/commit/6a...a442ec0a22
"Fix Any.end/elems"
FROGGS hmmm
13:51 domidumont left
masak that's when it changed from `self.list.elems` to `0` for the type object 13:51
13:51 araujo__ joined
masak I think it might be a thinko 13:51
13:51 Sqirrel left
jnthn quite possibly, yes 13:51
FROGGS I'll spectest a change
masak something like "well, the type object has no elements"
[Coke] colomon: (counting) yes, as long as it has a plan.
13:51 domidumont joined
tadzik FROGGS: is github.com/tadzik/panda/issues/112 fixed? 13:51
13:52 araujo_ left
TimToady I think that was an attempt to fix index() before we had with/without 13:52
colomon [Coke]++
13:53 araujo__ left
dalek ast/glr: 93eff8e | (Stefan Seifert)++ | S16-io/say.t:
Tests for GLRified semantics of say
 13:53
kudo/glr: bb736ad | (Stefan Seifert)++ | src/core/io_operators.pm:
Fix say @array not iterating over the arguments items
masak the GLRification of say 13:54
[Coke] r: use MONKEY_TYPING; augment slang Regex { token numish { \d+ } };
GLRelia rakudo-moar 965539: OUTPUT«===SORRY!===␤QAST::Block with cuid cuid_1_1440683686.69991 has not appeared␤Saw 1 occurrence of deprecated code.␤================================================================================␤'use MONKEY_TYPING' seen at:␤ /tmp/tmpfile, line 1…»
..rakudo-jvm 965539: OUTPUT«Can't open perl script "/home/steve/rakudo-inst/bin/eval-client.pl": No such file or directory␤»
camelia rakudo-moar 47ddca: OUTPUT«===SORRY!===␤QAST::Block with cuid cuid_1_1440683686.63467 has not appeared␤Saw 1 occurrence of deprecated code.␤================================================================================␤'use MONKEY_TYPING' seen at:␤ /tmp/tmpfile, line 1…»
..rakudo-jvm 47ddca: OUTPUT«===SORRY!===␤setcodeobj can only be used with a CodeRef␤Unhandled exception: java.lang.NullPointerException␤ in print_exception (gen/jvm/CORE.setting:15481)␤ in <anon> (gen/jvm/CORE.setting:15522)␤ in command_eval (gen/jvm/stage2/NQPHLL.nqp:130…»
masak whoa 13:55
FROGGS r: use MONKEY-TYPING; augment slang Regex { token numish { \d+ } };
GLRelia rakudo-jvm 965539: OUTPUT«Can't open perl script "/home/steve/rakudo-inst/bin/eval-client.pl": No such file or directory␤»
..rakudo-moar 965539: OUTPUT«===SORRY!===␤QAST::Block with cuid cuid_1_1440683724.15141 has not appeared␤»
camelia rakudo-jvm 47ddca: OUTPUT«===SORRY!===␤setcodeobj can only be used with a CodeRef␤»
..rakudo-moar 47ddca: OUTPUT«===SORRY!===␤QAST::Block with cuid cuid_1_1440683724.08868 has not appeared␤»
masak whatever that was, it hit *both* Moar *and* JVM
[Coke]: could you please file that one?
[Coke] m: my Str constant foo = 1 ; 13:56
camelia ( no output )
GLRelia ( no output )
[Coke] masak: It's already filed.
FROGGS [Coke]: you're checking old tickets?
[Coke] FROGGS: yes.
13:56 jkva left
FROGGS [Coke]++ 13:56
masak [Coke]: the 'QAST::Block with cuid cuid_1_1440683686.69991 has not appeared' thing?
[Coke]: because that seems it should never ever happen. 13:57
tadzik FROGGS: is there something in existence that parses JSON::Slow:auth<cpan:*>:ver(1..*)-like things? 14:00
[Coke] r: gist.github.com/diakopter/2830046
GLRelia rakudo-moar bb736a: OUTPUT«5===SORRY!5=== Error while compiling /tmp/tmpfile␤Shaped variable declarations not yet implemented. Sorry. ␤at /tmp/tmpfile:1␤------> 3my @d[4;4]7⏏5;␤»
..rakudo-jvm bb736a: OUTPUT«Can't open perl script "/home/steve/rakudo-inst/bin/eval-client.pl": No such file or directory␤»
camelia rakudo-{moar,jvm} 47ddca: OUTPUT«5===SORRY!5=== Error while compiling /tmp/tmpfile␤Shaped variable declarations not yet implemented. Sorry. ␤at /tmp/tmpfile:1␤------> 3my @d[4;4]7⏏5;␤»
[Coke] masak: yes, that's ticketed.
14:00 breinbaas joined
TimToady m: say [[1,2,3],<a b c>].perl 14:01
camelia rakudo-moar 47ddca: OUTPUT«[[1, 2, 3], "a", "b", "c"]␤»
GLRelia rakudo-moar bb736a: OUTPUT«[[1, 2, 3], ("a", "b", "c")]␤»
masak [Coke]: ok, excellent.
[Coke] rt.perl.org/Ticket/Display.html?id=117193 - opened by masak. 14:02
masak don't know who that is 14:03
but great that there's a ticket
[Coke] m: say (1, 2, *) Z (1, 2, 3, *) 14:05
camelia rakudo-moar 47ddca: OUTPUT«1 1 2 2 * 3␤»
GLRelia rakudo-moar bb736a: OUTPUT«1 12 2* 3␤»
lizmat ｢ok 7 - Match.note 14:08
｣
14:09 aborazmeh left
FROGGS tadzik: ask the one to your left... when he is back 14:10
tadzik ok :)
dalek ast: 9208bf0 | coke++ | S32-io/IO-Socket-Async.t:
This test passing on mac os x now
 14:12
14:12 itz joined
[Coke] m: (1..*).eager 14:14
camelia ( no output )
14:14 abaugher left
GLRelia rakudo-moar bb736a: OUTPUT«(timeout)» 14:14
FROGGS star-m: my $a; say $a.elems
14:14 GLRelia left
camelia star-m 2015.03: OUTPUT«0␤» 14:14
FROGGS m: my $a; say $a.elems
camelia rakudo-moar 47ddca: OUTPUT«0␤»
14:15 abaugher joined
laben m: (1..*).eager.say 14:15
camelia rakudo-moar 47ddca: OUTPUT«1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 8…»
14:15 GLRelia joined
FROGGS m: my $a; say $a.elems 14:15
camelia rakudo-moar 47ddca: OUTPUT«0␤»
GLRelia star-m : OUTPUT«Can't chdir to '/home/steve/star/': No such file or directory at lib/EvalbotExecuter.pm line 171.␤ EvalbotExecuter::_auto_execute(HASH(0x2070390), "my \$a; say \$a.elems", GLOB(0x23e3288), "/tmp/Mm28zjppdC", "star-m") called at lib/EvalbotExecuter.pm line 119␤ Ev…»
rakudo-moar bb736a: OUTPUT«0␤» 14:16
colomon m: [ 0 xx 10, 10 xx 10, 20 xx 10 ].perl.say
camelia rakudo-moar 47ddca: OUTPUT«[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20]␤»
GLRelia rakudo-moar bb736a: OUTPUT«[(0, 0, 0, 0, 0, 0, 0, 0, 0, 0), (10, 10, 10, 10, 10, 10, 10, 10, 10, 10), (20, 20, 20, 20, 20, 20, 20, 20, 20, 20)]␤»
[Coke] r: say &infix:<max>(|(1..*))
camelia rakudo-{moar,jvm} 47ddca: OUTPUT«-Inf␤»
GLRelia rakudo-moar bb736a: OUTPUT«(timeout)»
..rakudo-jvm bb736a: OUTPUT«Can't open perl script "/home/steve/rakudo-inst/bin/eval-client.pl": No such file or directory␤»
dalek ast/glr: 2c53e61 | lizmat++ | S16-io/note.t:
Make note test work
tadzik timotimo++ # JSON::Fast
colomon Any easy way to get the flat version back in GLR? of [ 0 xx 10, 10 xx 10, 20 xx 10 ] 14:17
FROGGS m: my $a; say $a.elems # jnthn: this will also report '1' with my patch
nine m: say [flat 0 xx 10, 10 xx 10, 20 xx 10 ].perl
camelia rakudo-moar 47ddca: OUTPUT«0␤»
GLRelia rakudo-moar bb736a: OUTPUT«0␤»
rakudo-moar bb736a: OUTPUT«[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20]␤»
camelia rakudo-moar 47ddca: OUTPUT«[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20]␤»
colomon nine++
[Coke] jnthn: can you verify that RT #122948 is closable given glr? I am pretty sure it is, but it's written using old ".infinite" syntax. the bug that it was the root cause of is already fixed in GLR... 14:19
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=122948
dalek kudo/glr: 2023f02 | TimToady++ | src/core/ (2 files):
try out gist bracketing
 14:22
TimToady ^^^ is likely to break a few spectests 14:23
[Coke] # XXX -- Lists are not real datatypes, but I haven't found a better location ---- This comment seems deletable. S02-types/list.t 14:24
colomon [Coke]: :)
14:25 laouji joined
dalek ast: edea297 | coke++ | S02-types/lists.t:
Remove very old comment.
 14:25
14:25 xzenfar left
[Coke] .ell-ems 14:26
dalek ast/glr: 28feb5b | colomon++ | S32-list/categorize.t:
Make $arrayer do the right thing under GLR.
timotimo gist.github.com/timo/f1324355a5ae8dc95d36 - this is what it looks like when you accidentally run rakudo/moar against roast/glr 14:27
muraiki m: say 0o7.WHAT 14:28
GLRelia rakudo-moar bb736a: OUTPUT«␤»
camelia rakudo-moar 47ddca: OUTPUT«(Int)␤»
[Coke] m: role sidecat is {} ; my @a = 1,2,3 but sidecat; say @a.gist
camelia rakudo-moar 47ddca: OUTPUT«5===SORRY!5=== Error while compiling /tmp/ewdrdM6NJv␤Invalid name␤at /tmp/ewdrdM6NJv:1␤------> 3role sidecat is7⏏5 {} ; my @a = 1,2,3 but sidecat; say @a.␤»
GLRelia rakudo-moar bb736a: OUTPUT«5===SORRY!5=== Error while compiling /tmp/PAaQsjh1c7␤Invalid name␤at /tmp/PAaQsjh1c7:1␤------> 3role sidecat is7⏏5 {} ; my @a = 1,2,3 but sidecat; say @a.␤»
muraiki hrm, is there any way to find out that it's octal? for instance, if I want to use it as a type in a sub parameter
[Coke] m: role sidecat {} ; my @a = 1,2,3 but sidecat; say @a.gist 14:29
camelia rakudo-moar 47ddca: OUTPUT«1 2 3␤»
GLRelia rakudo-moar bb736a: OUTPUT«1 2 3␤»
timotimo rakudo/nom is what i meant to write
[Coke] m: role sidecat {} ; my @a = 1,2,3 but sidecat; my @b; @b.pick(2); say @b.gist
camelia rakudo-moar 47ddca: OUTPUT«␤»
GLRelia rakudo-moar bb736a: OUTPUT«␤»
[Coke] m: role sidecat {} ; my @a = 1,2,3 ; my @b; @b.pick(2) but sidecat; say @b.gist
GLRelia rakudo-moar bb736a: OUTPUT«␤»
camelia rakudo-moar 47ddca: OUTPUT«␤»
timotimo muraiki: no, something that's octal is turned into an Int immediately
[Coke] m: role sidecat {} ; my @a = 1,2,3 ; my @b; @b = @a.pick(2) but sidecat; say @b.gist
GLRelia rakudo-moar bb736a: OUTPUT«3 2␤»
camelia rakudo-moar 47ddca: OUTPUT«This type does not support elems␤ in block <unit> at /tmp/EtQ8JFL7kd:1␤␤»
[Coke] m: role sidecat {} ; my @a = 1,2,3 ; my @b; @b = @a.pick(*).sort but sidecat; say @b.gist 14:30
camelia rakudo-moar 47ddca: OUTPUT«1 2 3␤»
GLRelia rakudo-moar bb736a: OUTPUT«1 2 3␤»
timotimo but how about <0o7>
m: say <0o7>.WHAT
camelia rakudo-moar 47ddca: OUTPUT«(Str)␤»
GLRelia rakudo-moar bb736a: OUTPUT«␤»
timotimo hm.
14:30 rindolf joined
dalek kudo/glr: ab4d41f | colomon++ | src/core/Hash.pm:
Get most of categorize working again on GLR.

This is somewhat hacky, definitely needs work on the more complicated case, probably could use some efficiency improvements to bring it back in line with the nom version.		 14:30
14:31 laben left
colomon that’s 27 additional passing tests for GLR 14:31
FROGGS \o/
timotimo way cool!
14:31 laben joined
colomon I’m not sure if the 27 were included in 81/80/77 counts tody, as categorize.t previously crashed before executing the first test. ;) 14:32
[Coke] m: role sidecat {} ; my @a = 1,2,3 ; my @b; @b = @a.pick(*).sort.list but sidecat; say @b.gist
camelia rakudo-moar 47ddca: OUTPUT«This type does not support elems␤ in block <unit> at /tmp/5agqRSwWlw:1␤␤»
GLRelia rakudo-moar 2023f0: OUTPUT«[1 2 3]␤»
[Coke] \o/
nine tests, not test files surely? 14:33
colomon oh, was that 81 test files failing? 14:34
jnthn colomon: The number has been test files, yes
nine Yes. But I think it was 76 last count
Running another one right now.
colomon jnthn: ah 14:35
well, then I’m not helping there at all, there is still one failing test
dalek kudo/glr: fffbe76 | lizmat++ | src/core/io_operators.pm:
Get note in line with say (with all the quirks)
lizmat this may introduce note breakage again
working on that
14:36 yqt joined
dalek ast/glr: 00cbae5 | coke++ | S02-types/list.t:
Add a test for RT #116527

only passes on glr		 14:36
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=116527
hahainternet is there a more succinct way to substr than .substr(0,10)? 14:38
[Coke] 1041 tickets.
we can go the CF route and add Left($_, 10) ! 14:39
muraiki timotimo: thanks!
jnthn hahainternet: No
hahainternet jnthn: merci 14:40
colomon oh!
TimToady m: $_ = "abcdefghijklmnopqrstuvwxyz"; say /^.**10/
camelia rakudo-moar 47ddca: OUTPUT«/^.**10/␤»
GLRelia rakudo-moar ab4d41: OUTPUT«/^.**10/␤»
TimToady m: $_ = "abcdefghijklmnopqrstuvwxyz"; say m/^.**10/
camelia rakudo-moar 47ddca: OUTPUT«｢abcdefghij｣␤»
GLRelia rakudo-moar ab4d41: OUTPUT«␤»
TimToady m: $_ = "abcdefghijklmnopqrstuvwxyz"; say ~m/^.**10/
GLRelia rakudo-moar ab4d41: OUTPUT«abcdefghij␤»
camelia rakudo-moar 47ddca: OUTPUT«abcdefghij␤»
TimToady that's more succinct :) 14:41
timotimo Visual Basic had "Mid" for substr (the general case)
jdv79 FROGGS: github.com/andk/pause/issues/182 - finally got it done; hopefully 14:42
ShimmerFairy I still think it'd be nice if substr used :end and :chars adverbs to specify the endpoint, so that it was less ambiguous in general 14:43
14:43 Ven left
ShimmerFairy needs to finish that S32::Stringy rewrite of hers soon... 14:43
dalek pan style="color: #395be5">perl6-examples: 98e0851 | (Sue Spence)++ | categories/cookbook/01strings/01-04converting-values.pl:
Change ord to ords, might as well convert the entire string
 14:44
pan style="color: #395be5">perl6-examples: a83f65d | (Sue Spence)++ | categories/cookbook/01strings/01-04converting-values.pl:
Fix comment
pan style="color: #395be5">perl6-examples: 14178a5 | (Sue Spence)++ | categories/cookbook/01strings/01-13upper-lower-case.pl:
Add example
pan style="color: #395be5">perl6-examples: 61f427f | (Sue Spence)++ | / (3 files):
Test fixes
pan style="color: #395be5">perl6-examples: 936f2a8 | (sue spence)++ | / (5 files):
Merge pull request #28 from virtualsue/master

More minor updates
[Coke] timotimo: CF has left, mid, and right. :|
kudo/glr: 7a4bd87 | colomon++ | src/core/Hash.pm:
Turns out the code I had commented out worked.

I was getting two different problems crossed up.
TimToady m: say [[1,2,3],<a b c>].gist
camelia rakudo-moar 47ddca: OUTPUT«1 2 3 a b c␤»
GLRelia rakudo-moar fffbe7: OUTPUT«[[1 2 3] (a b c)]␤»
colomon okay, now categorize.t completely passes.
hahainternet TimToady: that's quite cute :)
[Coke] m: sub foo { my $s = 0; ($s += $_ for 1..3) }; say foo() 14:45
GLRelia rakudo-moar fffbe7: OUTPUT«666␤»
camelia rakudo-moar 47ddca: OUTPUT«6 6 6␤»
hahainternet i was just thinking indexing into string or similar, i recognise that has ambiguity so *shrug*
i'm not a language designer
and i enjoy just writing what i can in p6 atm :)
[Coke] that's from rt.perl.org/Ticket/Display.html?id=77334 which we might want to revisit in the light of GLR.
timotimo [Coke]: i think that's actually the same as VB, then 14:46
TimToady m: sub foo { my $s = 0; (($s += $_)<> for 1..3) }; say foo()
camelia rakudo-moar 47ddca: OUTPUT«1 3 6␤»
GLRelia rakudo-moar fffbe7: OUTPUT«136␤»
dalek kudo/glr: 2003eac | colomon++ | src/core/Hash.pm:
Bring comments in line with the code again.
TimToady the problem is you're returning $s (the variable) three times
[Coke] m: my $a = 1; my $b = 2; $a = ($a, $b); say $a[0] 14:47
camelia rakudo-moar 47ddca: OUTPUT«Memory allocation failed; could not allocate 29952 bytes␤»
GLRelia rakudo-moar fffbe7: OUTPUT«(timeout)»
TimToady m: my $a = 1; my $b = 2; $a = (|$a, $b); say $a[0] 14:48
camelia rakudo-moar 47ddca: OUTPUT«5===SORRY!5=== Error while compiling /tmp/gc6pUtxwbw␤Arg-flattening | is only valid in an argument list␤at /tmp/gc6pUtxwbw:1␤------> 3my $a = 1; my $b = 2; $a = (|7⏏5$a, $b); say $a[0]␤»
GLRelia rakudo-moar fffbe7: OUTPUT«1␤»
[Coke] adding that to RT #82424
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...l?id=82424
FROGGS we are up to 85 spectests failing now, I'm about to fix the tests wrt parens outputted by .gist 14:50
[Coke] m: gist.github.com/coke/a19cab478ca24fb0a153 14:51
camelia rakudo-moar 47ddca: OUTPUT«5===SORRY!5=== Error while compiling /tmp/sORoDr0qBh␤Cannot use variable @hamming in declaration to initialize itself␤at /tmp/sORoDr0qBh:11␤------> 3my @hamming := (1, (@7⏏5hamming X* 2) merge (@hamming X* 3) merg␤ expecting any of:␤…»
GLRelia rakudo-moar 2003ea: OUTPUT«5===SORRY!5=== Error while compiling /tmp/KJVLyL4YdG␤Cannot use variable @hamming in declaration to initialize itself␤at /tmp/KJVLyL4YdG:11␤------> 3my @hamming := (1, (@7⏏5hamming X* 2) merge (@hamming X* 3) merg␤ expecting any of:␤…»
dalek c: d3ab2bf | paultcochrane++ | lib/Type/Exception.pod:
Document die and warn

  ... by shamelessly stealing from the specs text.		 14:53
14:53 donaldh left
dalek kudo/glr: fbe48b8 | FROGGS++ | src/core/Any-iterable-methods.pm:
add label as attribute to MapIterCommon

Before it was a lexical which just screamed BUG! jnthn++		 14:53
timotimo jnthn: in what places do i need to change stuff so that the statelist no longer ends up serialized, btw? 14:54
masak colomon++ # categorically passing
jnthn timotimo: NFA.nqp or so
timotimo and at what exact position do i put nfatostatelist back in?
nine 84 failing right now
jnthn In NQP
timotimo that's the file i'm looking at right now :)
jnthn oh
timotimo nfa_run and nfa_run_proto is where i've put it right now so that it exercises the nqp:: op
dalek kudo/glr: 5e98abb | jnthn++ | src/core/array_slice.pm:
Multi-dim array slicing (.[0,1;1]).
timotimo that's the place where the state list gets turned into an nfa if it hasn't been done yet 14:55
so perhaps after turning the statelist into the nfa object, the statelist ought to be cleared out?
jnthn timotimo: I'd do that and then see where it gets sad.
timotimo will do
jnthn timotimo: I thought we pre-computed the NFAs tough
*though
Look for precompute or so in Curor.nqp maybe 14:56
timotimo ah, indeed
nine m: say (1 => 'a', 2 => 'b') 14:57
GLRelia rakudo-moar 2003ea: OUTPUT«1 => a2 => b␤»
camelia rakudo-moar 47ddca: OUTPUT«1 => a 2 => b␤»
14:58 smls joined
timotimo so perhaps at the end of optimize would be a possible position where we could clear out the states list 14:59
jnthn perhaps so
14:59 Ven joined
timotimo sadly, this would come with a stage0 upgrade to get the nqp::nfatostatelist op available soon enough :| 15:00
dalek ast/glr: fdc2691 | (Stefan Seifert)++ | S02-types/pair.t:
Fix fallout from rakudo commit 2023f0218f19cd1a625f5a5b6ffb69a532db1053
timotimo but maybe we can do the stage0 update after we get the size improvements from no longer serializing the NFA's statelists 15:01
so it'd be a tiny bit less painful
dalek ast/glr: e0dc90d | (Stefan Seifert)++ | S03-metaops/reduce.t:
More fallout from rakudo commit 2023f0218f19cd1a625f5a5b6ffb69a532db1053
jnthn timotimo: Do it in a branch, then do the two, do the two rebootstraps, then rebase -i ;) 15:02
dalek ast/glr: a64915a | (Stefan Seifert)++ | integration/advent2013-day08.t:
More fallout from rakudo commit 2023f0218f19cd1a625f5a5b6ffb69a532db1053
colomon lizmat: I’m thinking that the multi-layer classify just doesn’t make sense for Baggy 15:03
dalek ast/glr: c53ca4f | FROGGS++ | S32-io/note.t:
More fallout from rakudo commit 2023f0218f19cd1a625f5a5b6ffb69a532db1053
timotimo ah, !protoregex wants to have the list of fates, but we can just keep only that around, perhaps in @!fates 15:04
so far it has been in $!states[0]
dalek ast/glr: 4430626 | (Stefan Seifert)++ | integration/advent2013-day15.t:
More fallout from rakudo commit 2023f0218f19cd1a625f5a5b6ffb69a532db1053
[Coke] colomon: is it expected that ./smoker projects.json would run for 40+ minutes at 100% cpu? 15:05
dalek kudo/glr: fb323f4 | jnthn++ | src/core/array_slice.pm:
Make * work in multi-dim slices.
timotimo or we could have a nqp::getnfafates op that pulls it out of the NFA object
colomon [Coke]: think so
timotimo that's basically a pointer steal
colomon [Coke]: well, it takes a long time, anyway.
timotimo no copying involved, and so we don't need to have it duplicated in memory either
colomon of course, it could be wedged, too.
… actually, that’s kind of weird 15:06
masak could someone show a use case of the multi-layer classify? I'm not sure I get it.
dalek ast/glr: 6fa9955 | (Stefan Seifert)++ | S12-class/mro.t:
More fallout from rakudo commit 2023f0218f19cd1a625f5a5b6ffb69a532db1053
 15:07
kudo/glr: c563ecd | lizmat++ | src/core/IO/Handle.pm:
Add a str candidate for IO::Handle.print
colomon [Coke]: just killed it and started over in a terminal, so I can watch what it’s doing
dalek ast/glr: 6faf49f | FROGGS++ | S32-io/pipe.t:
correctly say multiple lines
 15:08
skids o/
yoleaux 06:21Z <nine> skids: Ah, yes thanks! I'll have a look at it after breakfast. Now that I understand the problem it solves, I think it's a shame, your PR has not been merged yet.
06:39Z <nine> skids: thank you for fixing @a[1, *-1]:delete :) I've had a look at this yesterday but my train arrived before I came to writing code. I really wonder how this worked in nom though. Couldn't find any code handling that case there.
skids nine: np glad to be of help. Also, glad it wasn't in the slice code :-) 15:09
nine skids: btw. jnthn++ +1ed IterationBuffer.push not returning the pushed value.
skids Oh nice. 15:10
15:10 diana_olhovik_ left 15:14 larion left
dalek rl6-roast-data: f357b2c | coke++ | / (10 files):
today (automated commit)
 15:15
15:15 domidumont left
colomon m: reduce( -> $a, $b { $a.{$b}; }, {a => {b => {c => 42}}}, <a b c>).perl.say 15:16
GLRelia rakudo-moar fb323f: OUTPUT«42␤»
camelia rakudo-moar 47ddca: OUTPUT«42␤»
colomon m: reduce( -> $a, $b { $a.WHAT.say; $a.{$b}; }, {a => {b => {c => 42}}}, <a b c>).perl.say
GLRelia rakudo-moar c563ec: OUTPUT«␤␤␤42␤»
camelia rakudo-moar 47ddca: OUTPUT«(Hash)␤(Hash)␤(Hash)␤42␤»
colomon m: (a => 42).WHAT.say 15:17
camelia rakudo-moar 47ddca: OUTPUT«(Pair)␤»
GLRelia rakudo-moar c563ec: OUTPUT«␤»
colomon umm....
m: (a => 42)[0].WHAT.say
GLRelia rakudo-moar c563ec: OUTPUT«␤»
camelia rakudo-moar 47ddca: OUTPUT«(Pair)␤»
jnthn That looks wrong
15:18 rindolf left
colomon it’s not a Pair, it’s something without a name containing a Pair? 15:18
nine m: (a => 2)[0].^name.say; 15:19
camelia rakudo-moar 47ddca: OUTPUT«Pair␤»
GLRelia rakudo-moar c563ec: OUTPUT«Pair␤»
nine It seems to be just WHAT that's strange
colomon oh, never mind on that, got confused by the camelias switching places
m: say Pair.WHAT
camelia rakudo-moar 47ddca: OUTPUT«(Pair)␤»
GLRelia rakudo-moar c563ec: OUTPUT«␤»
nine m: (a => 2)[0].WHAT.^name.say
camelia rakudo-moar 47ddca: OUTPUT«Pair␤»
GLRelia rakudo-moar c563ec: OUTPUT«Pair␤»
jnthn Normally it's 'cus there's a .gist that isn't constrained by :D 15:20
colomon m: say %h; say %h.WHAT
camelia rakudo-moar 47ddca: OUTPUT«5===SORRY!5=== Error while compiling /tmp/2N_f_st9RL␤Variable '%h' is not declared␤at /tmp/2N_f_st9RL:1␤------> 3say 7⏏5%h; say %h.WHAT␤»
GLRelia rakudo-moar c563ec: OUTPUT«5===SORRY!5=== Error while compiling /tmp/TFKpWDbhdh␤Variable '%h' is not declared␤at /tmp/TFKpWDbhdh:1␤------> 3say 7⏏5%h; say %h.WHAT␤»
colomon m: my %h; say %h.WHAT
GLRelia rakudo-moar c563ec: OUTPUT«␤»
camelia rakudo-moar 47ddca: OUTPUT«(Hash)␤»
dalek pan style="color: #395be5">perl6-examples: d8073d1 | (Steve Mynott)++ | t/categories/cookbook/09directories.t:
work around changed behaviour of File::Temp
 15:21
pan style="color: #395be5">perl6-examples: a39a3a6 | (Steve Mynott)++ | t/000-check-dependencies.t:
fix number of tests by counting array
15:21 brrt joined 15:22 pmurias joined
[Coke] m: my $foo="+";&infix:($foo)(3,4).say 15:22
camelia rakudo-moar 47ddca: OUTPUT«5===SORRY!5=== Error while compiling /tmp/ZvKoN5SVAo␤You can't adverb ␤at /tmp/ZvKoN5SVAo:1␤------> 3my $foo="+";&infix:($foo)7⏏5(3,4).say␤»
GLRelia rakudo-moar c563ec: OUTPUT«5===SORRY!5=== Error while compiling /tmp/H59FHLSUEQ␤You can't adverb ␤at /tmp/H59FHLSUEQ:1␤------> 3my $foo="+";&infix:($foo)7⏏5(3,4).say␤»
15:22 pmurias_ joined 15:23 domidumont joined
FROGGS m: my $a = <stringify args>; note $a 15:23
GLRelia rakudo-moar c563ec: OUTPUT«stringifyargs␤»
camelia rakudo-moar 47ddca: OUTPUT«stringify args␤»
nine Fix for that coming up 15:25
for 1.WHAT.say that is
m: 1.WHAT.say
camelia rakudo-moar 47ddca: OUTPUT«(Int)␤»
GLRelia rakudo-moar c563ec: OUTPUT«␤»
dalek c: cdd3c1d | paultcochrane++ | lib/Language/modules.pod:
Add information about basic module structure and file extensions
 15:26
kudo/glr: 814d421 | jnthn++ | src/core/array_slice.pm:
Implement *-1 style things in multi-dim.
colomon nine: yeah, that seems wrongish. :)
15:26 ribasushi left
smls Oh, I see jnthn is back to multidim work? 15:26
jnthn smls: Well, re-instating @a[1;1] style things 15:27
nine smls: yep, he just didn't find something else to do ;)
JimmyZ seq.t would be good one
dalek kudo/glr: 3deda75 | (Stefan Seifert)++ | src/core/io_operators.pm:
Fix 1.WHAT.say not printing anything
 15:28
15:28 zakharyas left
nine m: my ($x, @b) = ('xxx', (1,2,3).map(*.Int)); say @b.perl; 15:29
camelia rakudo-moar 47ddca: OUTPUT«[1, 2, 3]<>␤»
GLRelia rakudo-moar c563ec: OUTPUT«[(1, 2, 3).Seq,]␤»
nine jnthn: any hint on where to look for that ^^^ would be appreciated
dalek ast/glr: bc14686 | (Stefan Seifert)++ | S16-io/say-and-ref.t:
Fix fallout from GLRification of say
 15:30
15:31 rurban left, khw joined
FROGGS m: my @a = any(3, 4); my $ok = 0; my $iterations = 0; for @a -> $x { $ok = 1 if $x ~~ Junction; $iterations++; }; say $ok; say $iterations 15:32
GLRelia rakudo-moar 814d42: OUTPUT«0␤1␤»
camelia rakudo-moar 47ddca: OUTPUT«1␤1␤»
colomon nine: I don’t think that was actually the problem I’m seeing?
nine unofficial count: 74
dalek c: f3935b3 | paultcochrane++ | lib/Type/X/Syntax/AddCategorial/MissingSeparator.pod:
Add missing title to X::Syntax::AddCategorial::MissingSeparator
jnthn nine: LIST.STORE
15:32 rurban joined
jnthn nine: Uh, List.STORE 15:32
timotimo STOPE IT, LIST 15:33
JimmyZ m: my @b = (1,2,3).map(*.Int); say @b.perl;
camelia rakudo-moar 47ddca: OUTPUT«[1, 2, 3]<>␤»
GLRelia rakudo-moar 814d42: OUTPUT«[1, 2, 3]␤»
colomon verb [ no obj. ] (usu. as noun stoping)
(in mining) excavate a series of steps or layers in (the ground or rock).
• (as noun stoping) Geology the process by which country rock is broken up and removed by the upward movement of magma.
(sorry, that’s stope). I had no idea.
masak jnthn: "Unhandled exception: ctxlexpad needs an MVMContext" -- what does this mean? :)
FROGGS m: my @a = any(3, 4); my $ok = 0; for @a -> $x { say 'loop'; say $x } 15:34
GLRelia rakudo-moar 814d42: OUTPUT«loop␤4␤»
camelia rakudo-moar 47ddca: OUTPUT«loop␤any(3, 4)␤»
Ven
.oO( it means you're masak`grr now )
15:34 laouji left
FROGGS m: my @a = any(3, 4); say @a 15:34
GLRelia rakudo-moar 814d42: OUTPUT«4␤»
camelia rakudo-moar 47ddca: OUTPUT«any(3, 4)␤»
jnthn masak: You used nqp::ctxlexpad on something that was't what it wanted
masak ok. 15:35
FROGGS masak: pass it nqp::ctx for example
timotimo it's possible that masak didn't use that himself :)
FROGGS m: nqp::ctxlexpad(nqp::ctx)
GLRelia rakudo-moar 814d42: OUTPUT«===============================================================================␤The use of nqp::operations has been deprecated for non-CORE code. Please␤change your code to not use these non-portable functions. If you really want␤to keep using nqp:…»
camelia rakudo-moar 47ddca: OUTPUT«===============================================================================␤The use of nqp::operations has been deprecated for non-CORE code. Please␤change your code to not use these non-portable functions. If you really want␤to keep using nqp:…»
masak actually all I did was run perl6-debug-m on my code :)
itz try turning it off and on again?
FROGGS m: use nqp; nqp::ctxlexpad(nqp::ctx)
camelia ( no output )
GLRelia ( no output )
Ven invokes ./masak-debug-m
colomon m: my @a; @a.push(Hash); say @a.perl
GLRelia rakudo-moar 814d42: OUTPUT«[]␤»
camelia rakudo-moar 47ddca: OUTPUT«[Hash]<>␤»
masak Ven: ctxlexpad! blblblbl!
colomon THAT’S what’s causing the problem I’m looking at.
masak ok, Grammar::Debugger also dies with a random unrelated error... :) 15:36
conclusion: my problem is too difficult for debugging tools!
"figure it out yourself, masak!" :P
Ven
.oO( who debugs the debugger? )
15:36 araujo joined, araujo left
masak .oO( qui debuggiet ipsos debuggerem? ) 15:36
15:37 araujo joined
[Coke] Ven: Pretty sure that's jnthn. 15:38
15:38 araujo left
colomon I think I’d doing someone wrong. People keep on going out and coming back with alcohol. I went out and bought water…. 15:39
dalek kudo/glr: 810daba | lizmat++ | src/core/Any.pm:
Add an .nl method so that we can concatenate .say
kudo/glr: 38d336d | lizmat++ | src/core/io_operators.pm:
Concatenate .say/.note/.print

The idea being that nqp::concat on MoarVM doesn't actually concatenate strings, but rather creates a linked list of the original strings in memory. So, concatenating strings this way should not create any memory pressure. It should however increase the change that different threads saying stuff, will actually not interleave their messages.
15:40 AlexDaniel joined
dalek kudo/glr: 15e124a | jnthn++ | src/core/array_slice.pm:
Fix .[*-2, *-1; 0].
 15:40
timotimo the circularity saw ... it bites :( 15:41
smls m: say (1, 2).WHAT
camelia rakudo-moar 47ddca: OUTPUT«(Parcel)␤»
GLRelia rakudo-moar 3deda7: OUTPUT«(List)␤»
smls hm, locally the GLR branch prints "Invocant requires an instance of type List, but a type object was passed." for that 15:42
masak timotimo: the fact that the circularity saw is *sharp* is probably the least fascinating fact about it :P
dalek ast/glr: 771f831 | jnthn++ | S09-subscript/multidim-assignment.t:
Update test for GLR.
 15:43
timotimo i'm not so successful with my endeavour
FROGGS m: my @a = 1 .. 5; @a[0,1] = @a[1,0]; say @a[0,1].WHAT
camelia rakudo-moar 47ddca: OUTPUT«(Parcel)␤»
GLRelia rakudo-moar 3deda7: OUTPUT«(List)␤»
colomon m: my @a; @a.push(Hash); say @a.perl 15:44
GLRelia rakudo-moar 3deda7: OUTPUT«[]␤»
camelia rakudo-moar 47ddca: OUTPUT«[Hash]<>␤»
colomon ^^ is GLR correct here, or is this another push issue?
FROGGS m: my @a = 1 .. 5; @a[0,1] = @a[1,0]; say @a[0,1] # I think the List.STORE method tries to be too cheat here too much
GLRelia rakudo-moar 15e124: OUTPUT«22␤»
camelia rakudo-moar 47ddca: OUTPUT«2 1␤»
FROGGS err, proper english difficult is 15:45
m: my @a; @a.push(Hash); say @a.elems
GLRelia rakudo-moar 15e124: OUTPUT«0␤»
camelia rakudo-moar 47ddca: OUTPUT«1␤»
FROGGS colomon: that is the same issue as Int.elems being zero
colomon FROGGS: … okay, but is that a bug? 15:46
FROGGS I very much think so
colomon it’s intentional: 15:47
proto method elems(|) is nodal { * }
15:47 rindolf joined
colomon multi method elems(Any:U:) { 0 } 15:47
multi method elems(Any:D:) { self.list.elems }
FROGGS colomon: we discussed that earlier... that zero there is a thinko
colomon FROGGS: … okay… it should be a 1? 15:48
FROGGS yes
colomon seems like a very easy fix?
FROGGS but that will cause other problems
colomon oh?
FROGGS m: my $a; say $a.elems
camelia rakudo-moar 47ddca: OUTPUT«0␤»
GLRelia rakudo-moar 15e124: OUTPUT«0␤»
FROGGS this will be 1 then
which might be okay, but I can't judge on that 15:49
m: my $a; for $a { say 42 }
GLRelia ( no output )
camelia rakudo-moar 47ddca: OUTPUT«42␤»
FROGGS yeah, it potentially should output 1 in the script before the last 15:50
dalek rl6-roast-data: 4dd10c4 | coke++ | / (2 files):
another glr run for 'today'
[Coke] can now enter his glr-rebuild-retest loop for the day, whee. 15:51
FROGGS colomon: also change method end
colomon FROGGS: to 0? 15:52
15:52 rindolf left
FROGGS aye 15:52
colomon those changes are old? lizmat++ made them in Jan? 15:53
FROGGS aye
colomon FROGGS: trying the changes to see what happens
FROGGS but pre-glr still iterated over things that had zero elems
lizmat what did I change in Jan ?
colomon lizmat: github.com/rakudo/rakudo/commit/6a...a442ec0a22
dalek kudo/glr: 3459537 | jnthn++ | src/core/ (2 files):
Implement {;} style indexing.
 15:54
colomon we’re pondering changing the values returned by the :U version of both of these
15:54 rindolf joined
lizmat to Nil or so? 15:54
[Coke] ... me kills and restarts to get jnthn++'s commit. :P
colomon m: my @a; say @a.elems
GLRelia rakudo-moar 15e124: OUTPUT«0␤»
camelia rakudo-moar 47ddca: OUTPUT«0␤»
colomon m: my @a; say @a.defined
camelia rakudo-moar 47ddca: OUTPUT«True␤»
GLRelia rakudo-moar 15e124: OUTPUT«True␤»
FROGGS lizmat: github.com/rakudo/rakudo/commit/6a...a442ec0a22
colomon hurm. 15:55
that didn’t fix the issue I’ve been looking at.
FROGGS :/
nine colomon: @a being empty after @a.push(Hash) is actually constitent with GLR semantics 15:56
colomon nine: you think?
nine FROGGS: ^^^
FROGGS nine: how do you push a typo object then?
skids Hash =:= {} though.
nine colomon: yes, because of the single-arg rule again. Hash is iterable, so it is flattened when being passed to push
skids erm !
colomon oh, maybe that’s why it didn’t fix my issue. 15:57
I think I’m just going to change the test…. ;0
nine Or...maybe it shouldn't because it's not actually a Hash but the type object for a Hash
skids Type objects probably should not be considered themselves iterable.
FROGGS nine: that's what I am thinking
[Coke] jnthn: github.com/coke/perl6-roast-data/b...ummary.out 15:58
Working on refreshing now.
timotimo m: say "Perl 6".ords 15:59
GLRelia rakudo-moar 345953: OUTPUT«80101114108823954␤»
camelia rakudo-moar 47ddca: OUTPUT«80 101 114 108 8239 54␤»
skids
.oO(unless they have an identity value that is iterable)
lizmat m: say "Perl 6".ords.item 16:00
GLRelia rakudo-moar 345953: OUTPUT«80101114108823954␤»
camelia rakudo-moar 47ddca: OUTPUT«80 101 114 108 8239 54␤»
dalek ast/glr: 1a0223c | colomon++ | S32-list/reduce.t:
Rework test to not push type objects.

Previously it pushed the $hash.WHAT, then checked whether or not it was Hash. Now it pushes the result of $hash ~~ Hash, and checks that it was true.		 16:01
colomon okay, that gets S32-list/reduce.t passing
16:02 rindolf left
[Coke] m: my @a := gather { for 1..3 { take $_; say @a.perl } }; say @a.perl 16:02
camelia rakudo-moar 47ddca: OUTPUT«()␤()␤()␤This type does not support positional operations␤ in block <unit> at /tmp/dKLzM_Yc_j:1␤␤»
GLRelia rakudo-moar 345953: OUTPUT«Type check failed in binding; expected 'Positional' but got 'Seq'␤ in block <unit> at /tmp/1dOVHEFQDc:1␤␤»
[Coke] ^^ RT #66820
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...l?id=66820
16:04 rindolf joined
lizmat m: sub a($a) { say $a.perl }; my %h = a => 42, b => 666; a %h 16:05
camelia rakudo-moar 47ddca: OUTPUT«{:a(42), :b(666)}␤»
GLRelia rakudo-moar 345953: OUTPUT«${:a(42), :b(666)}␤»
jnthn m: my %h = a => 1, b => 2; my @a = %h; say @a.perl 16:06
camelia rakudo-moar 47ddca: OUTPUT«[:a(1), :b(2)]<>␤»
GLRelia rakudo-moar 345953: OUTPUT«[:a(1), :b(2)]␤»
[Coke] m: say (^10).map: { $^n * 2 + 1 }.perl
camelia rakudo-moar 47ddca: OUTPUT«Cannot call map(Int: Int, Int, Int, Int, Int, Int, Int, Int, Int, Str); none of these signatures match:␤ ($: Whatever, *%_)␤ ($: &block, :$label, *%_)␤ in block <unit> at /tmp/W9wykctx3Y:1␤␤»
GLRelia rakudo-moar 345953: OUTPUT«Cannot call map(Range: Str); none of these signatures match:␤ ($: &block, :$label, *%_)␤ (HyperIterable:D $: &block, :$label, *%_)␤ in block <unit> at /tmp/ZWpV0OZm_I:1␤␤»
[ptc] Ven: stackoverflow.com/questions/1273616...atacompare
jnthn m: my %h = a => 1, b => 2; my @a = %h; say @a.map(*.WHAT).say
GLRelia rakudo-moar 345953: OUTPUT«Cannot call map(Array: Whatever); none of these signatures match:␤ ($: &block, :$label, *%_)␤ (HyperIterable:D $: &block, :$label, *%_)␤ in block <unit> at /tmp/OCTDnu6EC_:1␤␤»
camelia rakudo-moar 47ddca: OUTPUT«a => 1 b => 2␤True␤»
FROGGS m: my @a = gather { for 1..3 { take $_; say @a.perl } }; say @a.perl
GLRelia rakudo-moar 345953: OUTPUT«[]␤[]␤[]␤[1, 2, 3]␤»
camelia rakudo-moar 47ddca: OUTPUT«[]<>␤[]<>␤[]<>␤[1, 2, 3]<>␤»
lizmat m: sub a(\a) { say a.perl }; my %h = a => 42, b => 666; a %h
camelia rakudo-moar 47ddca: OUTPUT«{:a(42), :b(666)}<>␤»
GLRelia rakudo-moar 345953: OUTPUT«{:a(42), :b(666)}␤»
[ptc] Ven: as well as learn.perl.org/faq/perlfaq4.html#Ho...-are-equal
Ven [ptc]: thanks
lizmat m: sub a(**@a) { say @a.perl }; my %h = a => 42, b => 666; a %h 16:07
GLRelia rakudo-moar 345953: OUTPUT«[{:a(42), :b(666)},]␤»
camelia rakudo-moar 47ddca: OUTPUT«(:a(42), :b(666))␤»
lizmat m: sub a(*@a) { say @a.perl }; my %h = a => 42, b => 666; a %h
GLRelia rakudo-moar 345953: OUTPUT«[:a(42), :b(666)]␤»
camelia rakudo-moar 47ddca: OUTPUT«[:a(42), :b(666)]<>␤»
16:08 araujo joined, araujo left, araujo joined
lizmat m: sub a(*@a) { say @a.perl }; a { a => 42, b => 666 } 16:08
camelia rakudo-moar 47ddca: OUTPUT«[{:a(42), :b(666)}]<>␤»
GLRelia rakudo-moar 345953: OUTPUT«[:a(42), :b(666)]␤»
lizmat m: sub a(*@a) { say @a.perl }; a ${ a => 42, b => 666 }
GLRelia rakudo-moar 345953: OUTPUT«[{:a(42), :b(666)},]␤»
camelia rakudo-moar 47ddca: OUTPUT«5===SORRY!5=== Error while compiling /tmp/S_7n2s6Iq9␤Unsupported use of ${ a => 42, b => 666 }; in Perl 6 please use $( a => 42, b => 666 )␤at /tmp/S_7n2s6Iq9:1␤------> 3 say @a.perl }; a ${ a => 42, b => 666 }7⏏5<EOL>␤»
dalek kudo/glr: 9ba8e43 | TimToady++ | src/core/ (3 files):
1st whack at coping with self-referential data
 16:10
[Coke] m: sub foo(@a) { @a[1] = "NOES" }; my @a = <OH HAI>; foo(@a); say ~@a
camelia rakudo-moar 47ddca: OUTPUT«OH NOES␤»
GLRelia rakudo-moar 345953: OUTPUT«OH NOES␤»
[Coke] m: sub foo(@a) { @a[0, 1] = <OH NOES> }; my @a = 0, 1; foo(@a) 16:11
camelia ( no output )
GLRelia ( no output )
[Coke] m: sub foo(@a) { @a[0, 1] = <OH NOES> }; my @a = 0, 1; foo(@a); say ~@a;
camelia rakudo-moar 47ddca: OUTPUT«OH NOES␤»
GLRelia rakudo-moar 345953: OUTPUT«OH NOES␤»
FROGGS nine: what are you currently working on?
[Coke] masak: can you please review RT #74902? 16:12
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...l?id=74902
masak does
dalek ast/glr: 3fa8b4d | lizmat++ | S16-io/ (3 files):
note/print/say stupid test clean up
colomon m: rindex( 'xxy','y', -1e34 ) 16:13
camelia rakudo-moar 47ddca: OUTPUT«WARNINGS:␤Useless use of "rindex( 'xxy','y', -1e34 )" in expression "rindex( 'xxy','y', -1e34 )" in sink context (line 1)␤»
GLRelia rakudo-moar 345953: OUTPUT«WARNINGS:␤Useless use of "rindex( 'xxy','y', -1e34 )" in expression "rindex( 'xxy','y', -1e34 )" in sink context (line 1)␤»
colomon m: say rindex( 'xxy','y', -1e34 )
camelia rakudo-moar 47ddca: OUTPUT«Nil␤»
GLRelia rakudo-moar 345953: OUTPUT«Use of Nil in string context in block <unit> at /tmp/FgRhUOy8do:1␤␤»
masak [Coke]: no.
lizmat m: say rindex( 'xxy','y', -1 ) 16:14
GLRelia rakudo-moar 345953: OUTPUT«2␤»
camelia rakudo-moar 47ddca: OUTPUT«Nil␤»
masak [Coke]: I think you misunderstood the goal of the ticket.
nine FROGGS: I'm now changing everything, so itemization prevents from flattening under the single-arg-rule
[Coke] one of the goals was "these should be consistent"
but that wasn't the original request.
lizmat m: say $*FOO 16:15
camelia rakudo-moar 47ddca: OUTPUT«Dynamic variable $*FOO not found␤ in block <unit> at /tmp/xh0SL6RIGZ:1␤␤Actually thrown at:␤ in block <unit> at /tmp/xh0SL6RIGZ:1␤␤»
GLRelia rakudo-moar 345953: OUTPUT«Dynamic variable $*FOO not found␤ in block <unit> at /tmp/OV47SxHDTI:1␤␤Actually thrown at:␤ in block <unit> at /tmp/OV47SxHDTI:1␤␤»
[Coke] "so something is inconsistent, at least."
masak [Coke]: the spec said (and still *says*, I think) that the *contents* of the array parameter are readonly. that's what the "one level down" thing refers to.
[Coke]: I might try to find the relevant bit of spec for you.
dalek kudo/glr: 7467173 | TimToady++ | src/core/Array.pm:
remove useless dynvar declarations
 16:16
[Coke] please add it to the ticket if so, yes.
16:17 rangerprice joined
TimToady m: my $a = 1; my $b = 2; $a = ($a, $b); say $a[0] 16:18
GLRelia rakudo-moar 9ba8e4: OUTPUT«(\List_140492816507288 = (List_140492816507288 2))2␤»
camelia rakudo-moar 47ddca: OUTPUT«Memory allocation failed; could not allocate 29952 bytes␤»
dalek ast/glr: b057321 | FROGGS++ | S03-operators/repeat.t:
fix plan, was busted by 06b5ba0
 16:19
masak [Coke]: However, the default forces any container argument to also
be treated as an immutable value. This extends down only one level;
an immutable container may always return an element that is mutable if
it so chooses.
TimToady m: my $a = 1; my $b = 2; $a = (a => 2,:b($a, $b)); say $a.perl
GLRelia rakudo-moar 9ba8e4: OUTPUT«(my \List_140420932479392 = (:a(2), :b($(List_140420932479392, 2))))␤»
16:19 dwarring left
masak S06:718 16:19
synbot6 Link: design.perl6.org/S06.html#line_718
FROGGS m: my $a = 1; my $b = 2; $a = (a => 2,:b($a, $b)); say $a.perl.EVAL
GLRelia rakudo-moar 9ba8e4: OUTPUT«5===SORRY!5=== Error while compiling EVAL_0␤Variable '&List_140203337335216' is not declared␤at EVAL_0:1␤------> 3my \List_140203337335216 = (:a(2), :b($(7⏏5List_140203337335216, 2))))␤»
TimToady m: my $a = 1; my $b = 2; $a = {a => 2,:b($a, $b)}; say $a.perl
GLRelia rakudo-moar 9ba8e4: OUTPUT«(my \Hash_139742432439992 = {:a(2), :b($(Hash_139742432439992, 2))})␤»
masak adds this to the ticket 16:20
camelia rakudo-moar 47ddca: OUTPUT«(timeout)»
FROGGS m: my $a = 1; my $b = 2; $a = {a => 2,:b($a, $b)}; say $a.perl.EVAL
GLRelia rakudo-moar 9ba8e4: OUTPUT«5===SORRY!5===␤Variable '&Hash_140342462502600' is not declared␤at EVAL_0:1␤------> 3my \Hash_140342462502600 = {:a(2), :b($(7⏏5Hash_140342462502600, 2))})␤Variable '&Hash_140342462502600' is not declared␤at EVAL_0:1␤------> 3my \Has…»
FROGGS k, same 16:21
TimToady looks like my \foo isn't declaring the symbol soon enough
camelia rakudo-moar 47ddca: OUTPUT«(timeout)»
16:21 domidumont left
dalek p/nfa_to_statelist: e4f5004 | timotimo++ | src/QRegex/ (2 files):
try to nuke statelist and keep fates

but we're not getting far at all yet.		 16:21
TimToady std: (my \Hash_139742432439992 = {:a(2), :b($(Hash_139742432439992, 2))}) 16:22
16:22 GLRelia left
camelia std 28329a7: OUTPUT«ok 00:00 142m␤» 16:22
FROGGS TimToady: it does so potentially after the initializer?
TimToady std declares it soon enough :)
16:22 itz left, MilkmanDan left
TimToady m: (my \Hash_139742432439992 = {:a(2), :b($(Hash_139742432439992, 2))}) 16:23
camelia rakudo-moar 47ddca: OUTPUT«5===SORRY!5===␤Variable '&Hash_139742432439992' is not declared␤at /tmp/5tJjoJTYeL:1␤------> 3my \Hash_139742432439992 = {:a(2), :b($(7⏏5Hash_139742432439992, 2))})␤Variable '&Hash_139742432439992' is not declared␤at /tmp/5tJjoJTYeL:1…»
colomon how do we rebuild GLRelia?
16:23 GLRelia joined
GLRelia std : OUTPUT«Can't chdir to '/home/steve/std/snap': No such file or directory at lib/EvalbotExecuter.pm line 171.␤ EvalbotExecuter::_auto_execute(HASH(0x206fe38), "(my \\Hash_139742432439992 = {:a(2), :b(\$(Hash_139742432439992"..., GLOB(0x2417bc8), "/tmp/AKTCivk9Ni", "std") calle…» 16:23
colomon or is it only one commit behind? 16:24
masak ok, 007 now has custom operators \o/
colomon m: say rindex( 'xxy','y', -1e34 )
camelia rakudo-moar 47ddca: OUTPUT«Nil␤»
GLRelia rakudo-moar 746717: OUTPUT«Use of Nil in string context in block <unit> at /tmp/vdHBDNEDhb:1␤␤»
16:25 MilkmanDan joined
colomon is really confused about the difference he is getting between evalbot and his local p6 16:25
> say rindex( 'xxy','y', -1e34 ) 16:26
index start offset out of range
masak still have custom macro ops left to do...
[Coke] what's your local p6 version? 16:29
colomon [Coke]: 7467173fe344366f0c94f431dc42657c69d0f13c
masak that's GLR, it seems.
colomon yes
it’s the latest, so far as I know
I have been pulling regularly all afternoon
and GLRelia has been updated to it 16:30
moritz will shortly go off to celebrate his 5th wedding aniversary
brrt moritz++
colomon moritz++
FROGGS moritz: have fun! :o) 16:31
[Coke] moritz: congrats!
m: say (2015-1993)/(2015-1971) 16:32
camelia rakudo-moar 47ddca: OUTPUT«0.5␤»
GLRelia rakudo-moar 746717: OUTPUT«0.5␤»
moritz thanks everybody
hoelzro moritz: congrats!
timotimo moritz o/
jdv79 is there a terse way to coerce in a case like this?: gist.github.com/jdv/b92fa2a926d801804f83
FROGGS TimToady: the Hash_* is declared it the actions here: rakudo/src/Perl6/Actions.nqp:2435: elsif $<defterm> {
16:33 carlin left
moritz m: try { die 42 }; my $x = $!.WHAT; say $x; 16:33
camelia rakudo-moar 47ddca: OUTPUT«(X::AdHoc)␤»
GLRelia rakudo-moar 746717: OUTPUT«Use of uninitialized value $x of type X::AdHoc in string context in block <unit> at /tmp/CCpdXAmFhs:1␤␤»
16:33 carlin joined
[Coke] ^^ as of tomorrow, I'll have been alive only 2x as long as I've been married. Ish. 16:33
masak moritz: congrats
moritz [Coke]: wow.
[Coke] And I'm spending it with you guys!
masak aww :) 16:34
moritz m: try { die 42 }; my $x = $!.WHAT; say $x.gist;
camelia rakudo-moar 47ddca: OUTPUT«(X::AdHoc)␤»
GLRelia rakudo-moar 746717: OUTPUT«(X::AdHoc)␤»
tadzik moritz: congratulations to you both :)
moritz: have a lot of fun
moritz tadzik, masak, hoelzro: thanks
dalek rl6-roast-data: c74ed93 | coke++ | / (2 files):
another glr run for 'today'
 16:35
16:36 elohmrow left
[Coke] only 9 more passing tests. hurry up, people! ;) 16:36
16:37 virtualsue left
carlin has the panda precomp change caused anything submitted to testers.perl6 to have a build stage grade of "NA"? 16:37
[Coke] *furious investigation* 16:38
tadzik ...could have
pmurias_ timotimo: the nfa handling is getting changed again?
tadzik panda's been pretty aggresively mangled by me today 16:39
dalek kudo/glr: 355b999 | TimToady++ | src/core/Enum.pm:
handle self-referential pair as well in gist/perl
timotimo pmurias_: just an optimization
but i'm not doing well there at all
dalek ast/glr: 8256c7f | lizmat++ | S32-io/note.t:
Fix another note concat issue
 16:40
kudo/glr: 8d17aab | (Stefan Seifert)++ | src/core/Array.pm:
GLRR: itemization should prevent flattening single args

First part: my @arr = 1, 2; Array.new(@arr) and [@arr] return: [1, 2] Array.new($@arr) and [$@arr] return: [[1, 2]]		 16:41
FROGGS m: say (1, 2, 3).perl 16:42
camelia rakudo-moar 47ddca: OUTPUT«(1, 2, 3)␤»
GLRelia rakudo-moar 746717: OUTPUT«(1, 2, 3)␤»
colomon m: permutations(3).list.perl
GLRelia ( no output )
camelia ( no output )
colomon m: permutations(3).list.perl.say 16:43
camelia rakudo-moar 47ddca: OUTPUT«((0, 1, 2), (0, 2, 1), (1, 0, 2), (1, 2, 0), (2, 0, 1), (2, 1, 0))␤»
GLRelia rakudo-moar 746717: OUTPUT«((0, 1, 2), List_139685018653472, List_139685018725080, List_139685018769784, List_139685018841176, List_139685018885816)␤»
nine m: ((1,2,3),(2,3,4),(3,4,5)).perl.say
camelia rakudo-moar 47ddca: OUTPUT«((1, 2, 3), (2, 3, 4), (3, 4, 5))␤»
GLRelia rakudo-moar 746717: OUTPUT«((1, 2, 3), (2, 3, 4), (3, 4, 5))␤»
nine TimToady: ^^^ you got a bug in .perl. Also the reason for t/spec/S32-list/combinations.t failing 16:44
colomon oh, it’s .perl that is the issue? 16:45
m: permutations(3).list.say
camelia rakudo-moar 47ddca: OUTPUT«0 1 2 0 2 1 1 0 2 1 2 0 2 0 1 2 1 0␤»
GLRelia rakudo-moar 8d17aa: OUTPUT«(0 1 2)(0 2 1)(1 0 2)(1 2 0)(2 0 1)(2 1 0)␤»
pmurias_ timotimo: ok, I'm porting over the bitpacking from the CODEPOINT_I_LL and it's a bit annoying 16:47
16:47 pmurias_ left, pmurias left
timotimo ah, yeah 16:48
i can see that
dalek osystem: 483d972 | (John Scoles)++ | META.list:
Acme::Flutterby - An object-oriented interface to a butterfly. In what else but Perl 6.

This module provides a simplistic, but powerful, interface to a Butterfly.		 16:50
osystem: 1ab273a | FROGGS++ | META.list:
Merge pull request #54 from byterock/patch-1

Acme::Flutterby - An object-oriented interface to a butterfly. In wha…
tadzik hah! Along with Acme::Meow
[Coke] nqp::shell complains it requires 7 operands. Not in the docs. :P 16:52
16:53 brrt left
[Coke] (trying to write a test that will fail on the jvm) 16:53
16:53 dakkar left
dalek kudo/glr: 4050d34 | jnthn++ | src/core/Any.pm:
Bring back .tree.
 16:54
16:54 abraxxa left
FROGGS .oO( bring back, bring back, bring back the tree to me, to meeee! ) 16:55
maettu dinner @7:30 (outside) 16:56
tadzik \o/
maettu can already smell it
colomon m: permutations(3).list[0].WHAT.say 16:57
camelia rakudo-moar 47ddca: OUTPUT«(Parcel)␤»
GLRelia rakudo-moar 8d17aa: OUTPUT«Use of uninitialized value of type List in string context in block <unit> at /tmp/6U6qj8LGdf:1␤␤»
colomon m: permutations(3)[0].WHAT.say
camelia rakudo-moar 47ddca: OUTPUT«(Parcel)␤»
GLRelia rakudo-moar 8d17aa: OUTPUT«Use of uninitialized value of type List in string context in block <unit> at /tmp/5WEs8A9jds:1␤␤»
[Coke] maettu++ 16:58
masak m: sub x($/) { given "foo" { when /foo/ { say "yay" } } }; x(42) 16:59
camelia rakudo-moar 47ddca: OUTPUT«Cannot assign to a readonly variable or a value␤ in sub x at /tmp/3G_glWnQvA:1␤ in block <unit> at /tmp/3G_glWnQvA:1␤␤»
GLRelia rakudo-moar 4050d3: OUTPUT«Cannot assign to a readonly variable or a value␤ in sub x at /tmp/skf5wIwFrJ:1␤ in block <unit> at /tmp/skf5wIwFrJ:1␤␤»
16:59 carlin left
masak we could detect this statically. 16:59
dalek kudo/glr: 066bec0 | TimToady++ | src/core/ (4 files):
allow reuse of objects in self-ref constructs
masak if I ask "should we?", will someone pop in and say "no, that's too much complexity for too little gain" ? :)
TimToady colomon: that should fix your permutations vs perl bug 17:00
colomon TimToady: building now... 17:01
done
indeed!
TimToady++
masak background is this: in sufficiently advanced action methods, I tend to do some smartmatching.
colomon combinations still failing something
masak but I have to do it inside of a lexical sub which I call, completely for that purpose
17:01 Ven left
[Coke] FROGGS: how do I run a command and get at the output? 17:03
tadzik jvm: nqp::ordat('', 0)
m: 1 17:04
camelia rakudo-moar 47ddca: OUTPUT«WARNINGS:␤Useless use of constant integer 1 in sink context (line 1)␤»
GLRelia rakudo-moar 066bec: OUTPUT«WARNINGS:␤Useless use of constant integer 1 in sink context (line 1)␤»
tadzik j: nqp::ordat('', 0) 17:05
GLRelia rakudo-jvm 066bec: OUTPUT«Can't open perl script "/home/steve/rakudo-inst/bin/eval-client.pl": No such file or directory␤»
camelia rakudo-jvm 47ddca: OUTPUT«===SORRY!===␤java.lang.NullPointerException␤»
FROGGS [Coke]: in nqp?
dalek ast/glr: af9c772 | jnthn++ | S04-statements/for-scope.t:
Update test for [...] semantics.
 17:07
ast/glr: c1677a5 | jnthn++ | S06-signature/passing-arrays.t:
for $a { } with Iterable now iterates.
[Coke] updated question: how to do .perl in nqp?
FROGGS [Coke]: there is nothing like that IMO 17:08
dalek rl6-roast-data: 67f9a94 | coke++ | / (2 files):
another glr run for 'today'
[Coke] trying to diagnose why nqp::shell fails in the jvm; hard when there's no tests. :|
17:09 ribasushi joined
dalek ast/glr: 6401b9a | jnthn++ | S06-signature/sigilless.t:
This isn't a good way to test itemness any more.
 17:10
FROGGS TimToady: I tried to make @recursive.perl.EVAL work: gist.github.com/FROGGS/da2607dd8fc9d9774d6d 17:11
though, only partially successful
tadzik jvm broken! 17:12
j: nqp::null() 17:13
GLRelia rakudo-jvm 066bec: OUTPUT«Can't open perl script "/home/steve/rakudo-inst/bin/eval-client.pl": No such file or directory␤»
camelia rakudo-jvm 47ddca: OUTPUT«===SORRY!===␤java.lang.NullPointerException␤»
FROGGS colomon: how far did you get with the Any.elems patch? 17:15
colomon FROGGS: I aborted it in confusion. do you still reckon it’s the right thing to do? or should I just test it fully?
FROGGS: I’m multi-tasking a bit here… also trying to sort out smoker issues. :)
colomon tries building it again. 17:16
17:16 rindolf left
FROGGS colomon: I think it is the right thing todo, but I can pick up that task again 17:16
colomon: I'll do it 17:17
TimToady FROGGS: I was considering adding a <deftermnow> that reduces before parsing the initializer, then we could put it into Actions 17:19
17:19 domidumont joined, daxim joined, domidumont left
FROGGS TimToady: sounds like a plan 17:20
lizmat m: my %h{Any}; given a => 42 { say $_.key.WHICH; (%h.AT-KEY($_.key.WHICH) //= ($_.key => 0)).value += $_.value }; say %h.pairs # broken on GLR
camelia rakudo-moar 47ddca: OUTPUT«Str|a␤Str|a => a => 42␤»
GLRelia rakudo-moar 066bec: OUTPUT«Str|a␤Use of uninitialized value of type Any in numeric context in block <unit> at /tmp/Pvt0qV6JdZ:1␤Use of uninitialized value of type Any in numeric context in block <unit> at /tmp/Pvt0qV6JdZ:1␤ObjAt.new => :a(42)␤»
17:20 domidumont joined
jnthn m: my %h{Any}; given a => 42 { say $_.key.WHICH; (%h.AT-KEY($_.key.WHICH) //= ($_.key => 0)).perl.say }; 17:22
camelia rakudo-moar 47ddca: OUTPUT«Str|a␤:a(0)␤»
GLRelia rakudo-moar 066bec: OUTPUT«Str|a␤Use of uninitialized value of type Any in numeric context in block at /tmp/EYJQmrrKyq:1␤:a(0)␤»
jnthn m: my %h{Any}; given a => 42 { (($_.key => 0)).perl.say }; 17:23
camelia rakudo-moar 47ddca: OUTPUT«:a(0)␤»
GLRelia rakudo-moar 066bec: OUTPUT«Use of uninitialized value of type Any in numeric context in block at /tmp/mh9G0L58YP:1␤:a(0)␤»
jnthn m: given a => 42 { (($_.key => 0)).perl.say };
camelia rakudo-moar 47ddca: OUTPUT«:a(0)␤»
GLRelia rakudo-moar 066bec: OUTPUT«Use of uninitialized value of type Any in numeric context in block at /tmp/nLbZHXs7xa:1␤:a(0)␤»
jnthn m: given a => 42 { (($_.key => 0)).say };
camelia rakudo-moar 47ddca: OUTPUT«a => 0␤»
GLRelia rakudo-moar 066bec: OUTPUT«Use of uninitialized value of type Any in numeric context in block at /tmp/WgI9wOuUZF:1␤:a(0)␤»
jnthn m: given a => 42 { (($_.key => 0)).WHAT.say };
camelia rakudo-moar 47ddca: OUTPUT«(Pair)␤»
GLRelia rakudo-moar 066bec: OUTPUT«Use of uninitialized value of type Pair in string context in block at /tmp/4EtK1PD6fL:1␤␤»
17:24 espadrine left, rindolf joined
jnthn m: given a => 42 { (($_.key => 0)).DEFINITE }; 17:24
camelia ( no output )
GLRelia ( no output )
jnthn m: given a => 42 { (($_.key => 0)).DEFINITE.say };
camelia rakudo-moar 47ddca: OUTPUT«True␤»
GLRelia rakudo-moar 066bec: OUTPUT«True␤»
daxim hahainternet, what did you find re devanagari codepoints?
jnthn m: given a => 42 { (($_.key => 0)).^name.say };
camelia rakudo-moar 47ddca: OUTPUT«Pair␤»
GLRelia rakudo-moar 066bec: OUTPUT«Pair␤»
17:25 araujo left
jnthn m: given a => 42 { .key.say }; 17:25
camelia rakudo-moar 47ddca: OUTPUT«a␤»
GLRelia rakudo-moar 066bec: OUTPUT«a␤»
jnthn m: given a => 42 { (.key => 'x').say };
camelia rakudo-moar 47ddca: OUTPUT«a => x␤»
GLRelia rakudo-moar 066bec: OUTPUT«Use of uninitialized value of type Any in numeric context in block at /tmp/tAPfJOGESb:1␤:a("x")␤»
17:26 araujo joined, araujo left, araujo joined, araujo left
TimToady m: given a => 42 { (($_.key => 0)).WHICH.say } 17:26
camelia rakudo-moar 47ddca: OUTPUT«Pair|67325440␤»
GLRelia rakudo-moar 066bec: OUTPUT«Pair|52411280␤»
hahainternet daxim: that whatever is in rakudo as it stands doesn't work as the spec dictates 17:27
so i'll just wait
TimToady m: given a => 42 { (($_.key => 0)).WHAT }
camelia ( no output )
GLRelia ( no output )
hahainternet i'm not fussy, i appreciate there's a lot of work to be done
daxim I see 17:28
TimToady m: given a => 42 { (($_.key => 0)).WHAT.say }
camelia rakudo-moar 47ddca: OUTPUT«(Pair)␤»
GLRelia rakudo-moar 066bec: OUTPUT«Use of uninitialized value of type Pair in string context in block at /tmp/tWRmZ5JHer:1␤␤»
hahainternet i didn't bother anyone else here with it yet
m: "नि".chars
camelia ( no output )
GLRelia ( no output )
hahainternet m: say "नि".chars
camelia rakudo-moar 47ddca: OUTPUT«2␤»
GLRelia rakudo-moar 066bec: OUTPUT«2␤»
hahainternet so ya, if that doesn't work, nothing else will
but i know it's been on the list for ages so either spec or similar will be updated eventually 17:29
glr probably lot more important :D
lizmat m: my %b := BagHash.new(<a b o>,< p a p o o>); my %h = %b; say %b.perl
camelia rakudo-moar 47ddca: OUTPUT«Odd number of elements found where hash initializer expected␤ in block <unit> at /tmp/_McWAB6VDB:1␤␤»
GLRelia rakudo-moar 066bec: OUTPUT«("a"=>2,"p"=>2,"b"=>1,"o"=>3).BagHash␤»
jnthn m: say "नि".NFD.list.say 17:30
GLRelia rakudo-moar 066bec: OUTPUT«23442367␤True␤»
camelia rakudo-moar 47ddca: OUTPUT«2344 2367␤True␤»
jnthn m: say "नि".NFC.list.say
camelia rakudo-moar 47ddca: OUTPUT«2344 2367␤True␤»
GLRelia rakudo-moar 066bec: OUTPUT«23442367␤True␤»
jnthn m: say uniprop('Canonical_Combining_Class', 2367)
camelia rakudo-moar 47ddca: OUTPUT«Cannot call uniprop(Int, Int); none of these signatures match:␤ (Str $str, |c)␤ (Int $code, Stringy $propname = { ... })␤ in block <unit> at /tmp/muOYuduNGf:1␤␤»
GLRelia rakudo-moar 066bec: OUTPUT«Cannot call uniprop(Int, Int); none of these signatures match:␤ (Str $str, |c)␤ (Int $code, Stringy $propname = { ... })␤ in block <unit> at /tmp/InTi71kC3O:1␤␤»
jnthn m: say uniprop(2367, 'Canonical_Combining_Class') 17:31
camelia rakudo-moar 47ddca: OUTPUT«0␤»
GLRelia rakudo-moar 066bec: OUTPUT«0␤»
hahainternet jnthn: don't spend any time on it, it was just a trivial post on HN i noticed
so please don't divert yourself on my account :)
jnthn hahainternet: Yeah, but anyway that explains everything; things that are combiners are identified in the NFG algo by having a CCC > 0
hahainternet i gotta say i don't quite understand that, but fwiw, that character is literally taken from raw.githubusercontent.com/perl6/sp...nicode.pod 17:32
which indicates that .chars would return 1
jnthn Well, the doc pre-dates any actual NFG implementation work 17:34
So may need bringing in line with what actually got done.
hahainternet perhaps, there still seems to be some disagreement there but it's not with me, here is the HN comment relating to it: news.ycombinator.com/item?id=10130002
so one assumes there should be a way to get a count that = 1 grapheme, but buggered if i know where it'd be, thanks for elucidating jnthn 17:35
17:36 araujo joined, araujo left, araujo joined 17:37 araujo_ joined
FROGGS m: say "".split("").perl 17:38
GLRelia rakudo-moar 066bec: OUTPUT«Nil␤»
camelia rakudo-moar 47ddca: OUTPUT«Nil␤»
tony-o lizmat: Q about the recommendation manager in S22. is the recommendation manager referring to .candidates() in a CUR, or is the recommendation manager something further up the food chain that is expected query all CURs in the scope and combine the results of their .candidates
FROGGS m: say "".split("").elems
camelia rakudo-moar 47ddca: OUTPUT«0␤»
GLRelia rakudo-moar 066bec: OUTPUT«0␤»
17:38 Peter_R left
FROGGS m: say "".split("").elems; say Nil.elems 17:39
camelia rakudo-moar 47ddca: OUTPUT«0␤0␤»
GLRelia rakudo-moar 066bec: OUTPUT«0␤0␤»
nine jnthn: what's the difference between [,] and my &infix:<op> := &[,]; [op]?
17:39 araujo_ left
pink_mist precedence? 17:40
TimToady &[,] is always treated as binary
[,] will be a list associative operator
17:41 araujo left
nine so [,] will hit a different meta op implementation than [op]? 17:41
TimToady yup
nine That would explain: 17:42
m: my &infix:<op> := &[,]; say ([op] [1,2,3]).perl; say ([,] [1,2,3]).perl;
camelia rakudo-moar 47ddca: OUTPUT«([1, 2, 3],)␤([1, 2, 3],)␤»
GLRelia rakudo-moar 066bec: OUTPUT«((1, 2), 3)␤(1, 2, 3)␤»
TimToady m: my &infix:<op> := &infix:<,>; say ([op] [1,2,3]).perl; say ([,] [1,2,3]).perl;
masak TimToady: is there a way to refer to the list associative operator as a noun?
camelia rakudo-moar 47ddca: OUTPUT«([1, 2, 3],)␤([1, 2, 3],)␤»
GLRelia rakudo-moar 066bec: OUTPUT«((1, 2), 3)␤(1, 2, 3)␤»
lizmat dinner& 17:43
TimToady hmm, looks like the long form is also misbehaving
17:43 Peter_R joined
TimToady m: say &infix:<,>.prec 17:45
camelia rakudo-moar 47ddca: OUTPUT«assoc => list, prec => g=␤»
GLRelia rakudo-moar 066bec: OUTPUT«Use of uninitialized value of type Any in numeric context in block <unit> at /tmp/DH2Qt_UjM0:1␤Use of uninitialized value of type Any in numeric context in block <unit> at /tmp/DH2Qt_UjM0:1␤:prec("g="):assoc("list")␤»
TimToady well, it says it's list associative
muraiki does anyone have any advice for daemonizing a perl6 script? 17:49
hahainternet what platform? 17:50
muraiki ubuntu 14.04
hahainternet a simple systemd unit file would do it for you in the 'correct' ubuntu-y way
muraiki 14.04 is still on upstart, which I was giong to use, but I didn't know if there was perhaps a better way 17:51
17:51 araujo joined, araujo left, araujo joined
hahainternet i was /just/ about to ask that :) 17:51
17:51 yqt left, araujo left
muraiki hehe :) 17:51
hahainternet i realised 14.04 was a bit older than i thought
muraiki a lot happens in a 1.5 years!
hahainternet but yeah, there are lots of options but i prefer the 'simple' systemd approach
17:54 MilkmanDan left 18:05 kernel joined, kernel is now known as Guest81400, MilkmanDan joined 18:08 Guest6532 left 18:13 araujo joined 18:15 Sqirrel joined 18:16 arafatkamaal joined
arafatkamaal Hi 18:17
a noob question
$ perl6 > my @a = 1, 2; 1 2 > say @a.elems 2 > say @a[10] (Any) >
shouldn't this be an exception?
18:18 rurban left 18:19 cgfbee left
laben no, because it simply is not set. btw, if you assign, it automatically creates the memory slot. check autovivification 18:20
arafatkamaal i get it, but I was trying to access something that didn't exist 18:21
18:21 lucasb joined
ugexe m: my @a = 1, 2; 1 2 > say @a.elems 2 > say @a[10] 18:21
camelia rakudo-moar 47ddca: OUTPUT«5===SORRY!5=== Error while compiling /tmp/tMIwJwyIxj␤Two terms in a row␤at /tmp/tMIwJwyIxj:1␤------> 3my @a = 1, 2; 17⏏5 2 > say @a.elems 2 > say @a[10]␤ expecting any of:␤ infix␤ infix stopper␤ statement end…»
GLRelia rakudo-moar 066bec: OUTPUT«5===SORRY!5=== Error while compiling /tmp/rmyErt_y2J␤Two terms in a row␤at /tmp/rmyErt_y2J:1␤------> 3my @a = 1, 2; 17⏏5 2 > say @a.elems 2 > say @a[10]␤ expecting any of:␤ infix␤ infix stopper␤ statement end…»
ugexe or do you mean the (Any) > should throw an exception
laben m: my @a = 1, 2; 1 2; say @a.elems; say @a[10]
camelia rakudo-moar 47ddca: OUTPUT«5===SORRY!5=== Error while compiling /tmp/g5awu6DryG␤Two terms in a row␤at /tmp/g5awu6DryG:1␤------> 3my @a = 1, 2; 17⏏5 2; say @a.elems; say @a[10]␤ expecting any of:␤ infix␤ infix stopper␤ statement end␤ …»
GLRelia rakudo-moar 066bec: OUTPUT«5===SORRY!5=== Error while compiling /tmp/KFd66JBuWU␤Two terms in a row␤at /tmp/KFd66JBuWU:1␤------> 3my @a = 1, 2; 17⏏5 2; say @a.elems; say @a[10]␤ expecting any of:␤ infix␤ infix stopper␤ statement end␤ …»
laben m: my @a = 1, 2 say @a.elems; say @a[10]
camelia rakudo-moar 47ddca: OUTPUT«5===SORRY!5=== Error while compiling /tmp/ZZFn7SkwuF␤Two terms in a row␤at /tmp/ZZFn7SkwuF:1␤------> 3my @a = 1, 27⏏5 say @a.elems; say @a[10]␤ expecting any of:␤ infix␤ infix stopper␤ postfix␤ state…»
GLRelia rakudo-moar 066bec: OUTPUT«5===SORRY!5=== Error while compiling /tmp/Ef0b3Czhbs␤Two terms in a row␤at /tmp/Ef0b3Czhbs:1␤------> 3my @a = 1, 27⏏5 say @a.elems; say @a[10]␤ expecting any of:␤ infix␤ infix stopper␤ postfix␤ state…»
laben aargh
m: my @a = 1, 2; say @a.elems; say @a[10]
camelia rakudo-moar 47ddca: OUTPUT«2␤(Any)␤»
GLRelia rakudo-moar 066bec: OUTPUT«2␤Use of uninitialized value @a of type Any in string context in block <unit> at /tmp/LaW8Ks804c:1␤␤»
arafatkamaal That was copy pasted from the repl
18:22 cgfbee joined
geekosaur hm, LTA error from glr there 18:22
laben i think he means that @a[10] should throw exception since it doesnt exists
arafatkamaal yes
[Coke] last batch run is a mixed bag, some improvements, some new failures. 18:23
geekosaur perl6 is still perl; @a does not have a defined size so uninitialized values are undef ((Any) since it is also not typed)
18:23 llfourn left
arafatkamaal ok 18:23
laben arafatkamaal: it doesnt because of autovivification. since you did not declare a size for it (not a common case), it automatically expands
arafatkamaal so any list is basically of an unlimited size? 18:24
ugexe my @a[10] (shaped arrays) will do what you want but its not implemented yet i dont think
arafatkamaal ok
laben @a is an Array. it is not of unlimited size, but can be prolonged how much you need/want
pink_mist my @a = 1..Inf; #yes they are
[Coke] m: my @a=1..Inf; say @a.WHAT # pedant. 18:25
camelia rakudo-moar 47ddca: OUTPUT«(Array)␤»
GLRelia rakudo-moar 066bec: OUTPUT«Cannot look up attributes in a type object␤ in block <unit> at /tmp/TtrkN57RCw:1␤␤»
laben m: my @a = 1..5; say @a.elems; @a[20]; say @a.elems; @a[20] = 20; say @a.elems;
camelia rakudo-moar 47ddca: OUTPUT«5␤5␤21␤»
GLRelia rakudo-moar 066bec: OUTPUT«5␤5␤21␤»
laben see ^^^
arafatkamaal ok
[Coke] pedant check passed, thoughI don't think GLR is right there.
laben btw, why did panda switch to JSON::Fast? Now i need to actually fix that up for glr... 18:26
[Coke] 516 failing tests remain on glr. 18:27
pink_mist [Coke]: my "yes they are" was to arafatkamaal anyway, and he mentioned lists not arrays :P 18:28
arafatkamaal my @a = 1 .. Inf; 18:30
say @a.elems
should give Inf?
or just run forever
lucasb m: say class { method gist { 'oh' }; method Str { 'hai' } } 18:31
camelia rakudo-moar 47ddca: OUTPUT«oh␤»
GLRelia rakudo-moar 066bec: OUTPUT«hai␤»
rindolf pink_mist: hi, sup? 18:32
18:32 arnsholt left
pink_mist evening rindolf .. been darn tired all day 18:33
18:33 arnsholt joined
rindolf pink_mist: ah, that sucks. 18:34
pink_mist: do you know why?
pink_mist no :/
rindolf pink_mist: ah.
pink_mist: :-(
pink_mist: I'm going to hit the shower now.
Bye.
18:35 domidumont1 joined
TimToady m: my &infix:<op> := &infix:<,>; say &infix:<op>.prec 18:35
camelia rakudo-moar 47ddca: OUTPUT«assoc => list, prec => g=␤»
GLRelia rakudo-moar 066bec: OUTPUT«Use of uninitialized value of type Any in numeric context in block <unit> at /tmp/Vy0P9xFYSP:1␤Use of uninitialized value of type Any in numeric context in block <unit> at /tmp/Vy0P9xFYSP:1␤:prec("g="):assoc("list")␤»
18:37 domidumont left
[Coke] m: my @a=1..Inf; say @a.elems; # arafatkamaal 18:39
camelia rakudo-moar 47ddca: OUTPUT«Inf␤»
GLRelia rakudo-moar 066bec: OUTPUT«Cannot .elems a lazy list␤ in block <unit> at /tmp/3hE79y_dX3:1␤␤Actually thrown at:␤ in block <unit> at /tmp/3hE79y_dX3:1␤␤»
[Coke] GLR is the new way.
arafatkamaal on my rakudo installation it just runs forever, will install from the glr branch 18:40
what should happen if @b = Inf .. 1;
18:43 MilkmanDan left 18:44 Peter_R left 18:45 arafatkamaal left, MilkmanDan joined
dalek kudo/glr: c5086d7 | TimToady++ | src/core/Enum.pm:
fix Pair.perl uninit bug
 18:46
[Coke] you can't sequence backwards.
m: my @a = 3..1; say @a.perl;
GLRelia rakudo-moar 066bec: OUTPUT«[]␤»
camelia rakudo-moar 47ddca: OUTPUT«[]<>␤»
18:47 yqt joined
FROGGS m: use Test; my $should_ret_empty_list1 = sub { return; 1 }; is $should_ret_empty_list1().elems, 0, "our sub returned an empty list (1)"; 18:47
GLRelia rakudo-moar 066bec: OUTPUT«ok 1 - our sub returned an empty list (1)␤»
camelia rakudo-moar 47ddca: OUTPUT«ok 1 - our sub returned an empty list (1)␤»
FROGGS I think this test is bogus
18:47 Peter_R joined
lucasb m: sub f(|) { 42 }; say 2 &f 3 18:48
camelia rakudo-moar 47ddca: OUTPUT«all(2, 42)␤»
GLRelia rakudo-moar 066bec: OUTPUT«(timeout)» 18:49
lucasb ^^ I think this should be an error, right? It should not call f, and it has nothing to do with junctions
[Coke] lucasb: it's parsing the & as a junction combiner.
colomon [Coke]: you can’t range backwards. you can sequence however you want, if you ask nicely.
[Coke] whoops, you'd think I'd know the difference after 10 years. 18:50
TimToady m: use Test; my $should_ret_empty_list1 = sub { return slip; 1 }; is $should_ret_empty_list1().elems, 0, "our sub returned an empty list (1)";
GLRelia rakudo-moar 066bec: OUTPUT«ok 1 - our sub returned an empty list (1)␤»
camelia rakudo-moar 47ddca: OUTPUT«5===SORRY!5=== Error while compiling /tmp/CbcfF8mSbM␤Undeclared routine:␤ slip used at line 1. Did you mean 'flip', 'skip'?␤␤»
18:52 domidumont1 left
lucasb [Coke]: oh, understood. It is parsing as '2 & f(3)' 18:53
I was seeing '&f' as a single token and as expecting a TTIAR error 18:54
*was
18:59 muraiki left 19:04 Peter_R left 19:06 Peter_R joined
FROGGS hmpf 19:06
and now I don't know if return() should return Nil, Empty or Slip 19:07
$ perl6 -e 'say Empty.elems, Nil.elems, Slip.elems'
011 # locally
m: say Empty.elems, Nil.elems, Slip.elems
GLRelia rakudo-moar c5086d: OUTPUT«000␤»
camelia rakudo-moar 47ddca: OUTPUT«5===SORRY!5=== Error while compiling /tmp/MqfB4kAuRk␤Undeclared name:␤ Slip used at line 1␤␤»
FROGGS m: say Empty.elems, Nil.elems
camelia rakudo-moar 47ddca: OUTPUT«00␤»
GLRelia rakudo-moar c5086d: OUTPUT«00␤»
lucasb m: sub f {}; say @ = f 19:08
camelia rakudo-moar 47ddca: OUTPUT«(Any)␤»
GLRelia rakudo-moar c5086d: OUTPUT«(Any)␤»
FROGGS $ perl6 -e 'say Empty.elems, Nil.elems, Slip.new.elems'
010
lucasb oops, sorry
dalek kudo/glr: f3d060e | TimToady++ | src/core/Array.pm:
only add comma if single array elem is iterable
 19:10
ast/glr: 485af56 | FROGGS++ | S04-statements/return.t:
use non-deprecated isa-ok
rindolf FROGGS: hi! How was dinner yesterday? 19:13
FROGGS rindolf: I don't remember ó.ò 19:14
rindolf FROGGS: ah.
19:14 Ven joined
jdv79 rindolf: do you do any p6 stuff? 19:15
rindolf jdv79: a little - Project Euler/etc.
jdv79: at the perl6-examples repository.
timotimo no lol! no project euler!
laben i still am not sure of this one-arg rule, why does [{a => 1, b => 1}] transform into [a => 1, b=> 2] instead of {a => 1, b =>2} ? 19:16
rindolf timotimo: PE is love! PE is life! ;-)
timotimo ban he!
skids laben: {...} is the one-arg for [...]
FROGGS okay... now I understand that Empty is not much more than Slip.new 19:17
laben m: sub foo(\a) { dd a }; foo([{a => 1, b => 2}])
camelia rakudo-moar 47ddca: OUTPUT«$ = [{:a(1), :b(2)}]␤»
GLRelia rakudo-moar f3d060: OUTPUT«[:a(1), :b(2)]␤»
skids Is Empty even staying? 19:18
timotimo what's our current stance on how that's supposed to work? what laben just wrote?
with a dollar sign?
FROGGS skids: perhaps as Sugar...
skids I'm surmising from the "GLRR" commit that it's been decided you need to itemize. 19:19
FROGGS m: sub foo(\a) { dd a }; foo([${a => 1, b => 2}])
camelia rakudo-moar 47ddca: OUTPUT«5===SORRY!5=== Error while compiling /tmp/7gsHX1U9ZJ␤Unsupported use of ${a => 1, b => 2}; in Perl 6 please use $(a => 1, b => 2)␤at /tmp/7gsHX1U9ZJ:1␤------> 3foo(\a) { dd a }; foo([${a => 1, b => 2}7⏏5])␤»
GLRelia rakudo-moar f3d060: OUTPUT«[{:a(1), :b(2)},]␤»
skids m: sub foo(\a) { dd a }; foo($[${a => 1, b => 2}])
camelia rakudo-moar 47ddca: OUTPUT«5===SORRY!5=== Error while compiling /tmp/T6hLavoSNh␤Unsupported use of ${a => 1, b => 2}; in Perl 6 please use $(a => 1, b => 2)␤at /tmp/T6hLavoSNh:1␤------> 3oo(\a) { dd a }; foo($[${a => 1, b => 2}7⏏5])␤»
GLRelia rakudo-moar f3d060: OUTPUT«[{:a(1), :b(2)},]␤»
skids m: sub foo(\a) { dd a }; foo($[{a => 1, b => 2}]) 19:20
camelia rakudo-moar 47ddca: OUTPUT«$v = Any␤»
GLRelia rakudo-moar f3d060: OUTPUT«[:a(1), :b(2)]␤»
rangerprice i thing that Rust will be my preferred language after Perl
19:21 bin_005 joined
skids m: sub foo(\a) { dd a }; foo($[$[a => 1, b => 2]]) 19:22
camelia rakudo-moar 47ddca: OUTPUT«Method 'Int' not found for invocant of class 'Pair'␤ in block <unit> at /tmp/QezSscG3tw:1␤␤»
GLRelia rakudo-moar f3d060: OUTPUT«[[:a(1), :b(2)],]␤»
timotimo ich habe mir "birdemic" 1 & 2 empfehlen lassen 19:23
ww
laben m: sub foo(\a) { dd a }; foo([%{a => 1, b => 2}]) 19:25
GLRelia rakudo-moar f3d060: OUTPUT«[:a(1), :b(2)]␤»
camelia rakudo-moar 47ddca: OUTPUT«5===SORRY!5=== Error while compiling /tmp/WlWIJijYdI␤Unsupported use of %{a => 1, b => 2}; in Perl 6 please use %(a => 1, b => 2)␤at /tmp/WlWIJijYdI:1␤------> 3foo(\a) { dd a }; foo([%{a => 1, b => 2}7⏏5])␤»
laben m: sub foo(\a) { dd a }; foo([{a => 1, b => 2},])
camelia rakudo-moar 47ddca: OUTPUT«$ = [{:a(1), :b(2)}]␤»
GLRelia rakudo-moar f3d060: OUTPUT«[{:a(1), :b(2)},]␤»
tadzik soo, do we have a list of user-facing changes in glr?
like a porting guideline?
laben m: sub foo(\a) { dd a }; foo([{a => 1, b => 2},].elems) 19:26
camelia rakudo-moar 47ddca: OUTPUT«1␤»
GLRelia rakudo-moar f3d060: OUTPUT«1␤»
timotimo like a "Great Refactoring List"? 19:27
tadzik yes! 19:28
from another channel: [Perl is] a quantum superposition of everyone's favourite language
19:30 khw left
colomon jnthn: take @result.map(-> $x { $x }).list; 19:30
did NOT work. Same fail as before
laben timotimo: here is it a fix for JSON::Fast tests gist.github.com/6d4b4539b45368d70058
nine m: my @a = <a b c d e f>; say @a[[3, 4], 0,].perl; 19:31
camelia rakudo-moar 47ddca: OUTPUT«("c", "a")␤»
GLRelia rakudo-moar f3d060: OUTPUT«(("d", "e"), "a")␤»
nine m: my @a = <a b c d e f>; say @a[[3, 4]].perl; 19:32
GLRelia rakudo-moar f3d060: OUTPUT«("d", "e")␤»
camelia rakudo-moar 47ddca: OUTPUT«"c"␤»
nine m: my @a = <a b c d e f>; say @a[$[3, 4]].perl; 19:33
camelia rakudo-moar 47ddca: OUTPUT«Indexing requires an instance, tried to do: @a[ (Any) ]␤ in block <unit> at /tmp/bkatXDUy3b:1␤␤»
GLRelia rakudo-moar f3d060: OUTPUT«"c"␤»
laben m: my @a = <a b c d e f>; my @b = [3, 4]; say @a[@b].perl; 19:34
camelia rakudo-moar 47ddca: OUTPUT«("c",)␤»
GLRelia rakudo-moar f3d060: OUTPUT«("d", "e")␤»
laben m: my @a = <a b c d e f>; my @b = $[3, 4]; say @a[@b].perl;
camelia rakudo-moar 47ddca: OUTPUT«Indexing requires an instance, tried to do: @a[ (Any) ]␤ in block <unit> at /tmp/4CT76ieaNN:1␤␤»
GLRelia rakudo-moar f3d060: OUTPUT«("c")␤»
laben tadzik: panda/master is now the most updated branch, right? did you merge glr into master? 19:37
tadzik laben: I did not
was glr merged into rakudo master?
I mean, nom :)
laben: I was developing on nom all day
laben ok, so i will merge it in my repo manually
tadzik oh, thank you, that'd be sweet
laben i meant locally, or do you want me to push upstream? 19:38
dalek kudo/glr: ba5494a | TimToady++ | src/Perl6/ (2 files):
introduce my \term before initializer is parsed

  (Still doesn't work, since we haven't really closed the definitional loop yet.)
laben omfg a ton of conflicts, :( 19:39
tadzik :(
sorry
I may have changed whitespaces in places :(
totally didn't think of that 19:40
timotimo laben: is/isn't that going to be [${...}] in the future?
dalek ast/glr: 9081d01 | (Stefan Seifert)++ | S (2 files):
Adjust flatteing.t to new GLRR semantics

In a nutshell:
  * there is no implicit flattening
  * unless it's a container passed as single argument to a method or function
  * however, itemization prevents flattening even there
foo(@flattens) foo($@doesnt-flatten) foo(@we, @do, @no, @flatten)		 19:41
timotimo flat teing?
laben timotimo: ah is it? for now that does not work, even throws
timotimo mhh, i lost track :S
laben m: sub foo(\a) { dd a }; foo([$%{a => 1, b => 2}]) 19:42
GLRelia rakudo-moar f3d060: OUTPUT«[{:a(1), :b(2)},]␤»
camelia rakudo-moar 47ddca: OUTPUT«5===SORRY!5=== Error while compiling /tmp/9l0q3FLfS8␤Unsupported use of %{a => 1, b => 2}; in Perl 6 please use %(a => 1, b => 2)␤at /tmp/9l0q3FLfS8:1␤------> 3oo(\a) { dd a }; foo([$%{a => 1, b => 2}7⏏5])␤»
laben m: sub foo(\a) { dd a }; foo([${a => 1, b => 2}]) 19:43
camelia rakudo-moar 47ddca: OUTPUT«5===SORRY!5=== Error while compiling /tmp/KcB_WqMdsZ␤Unsupported use of ${a => 1, b => 2}; in Perl 6 please use $(a => 1, b => 2)␤at /tmp/KcB_WqMdsZ:1␤------> 3foo(\a) { dd a }; foo([${a => 1, b => 2}7⏏5])␤»
GLRelia rakudo-moar f3d060: OUTPUT«[{:a(1), :b(2)},]␤»
laben oh now it works
but cannot do, as it breaks nom-living perl6ers
19:44 virtualsue joined
TimToady m: my $a = 1; my $b = 2; $a = [1,2,($a, $b)]; say $a.perl.EVAL 19:46
camelia rakudo-moar 47ddca: OUTPUT«1 2 1 2␤»
GLRelia rakudo-moar ba5494: OUTPUT«12((Mu) 2)␤»
timotimo hm, so it displays it with a trailing comma ... maybe that's how it's going to stay? 19:47
TimToady progress, doesn't blow up, but we need a way of delaying evaluation of terms that are being defined until the binding
jdv79 masak: were you working on fixing eqv? i vaguely remember you were vocal about it. 19:48
TimToady timotimo: it's hard to do it otherwise
you end up with things like [$1..10]
skids GLRR++ now I know what the heck I'm supposed to be doing :-)
timotimo OK, if that's how it's going to stay, i'll merge laben's pull request for JSON::Fast
19:49 Peter_R left
TimToady trailing comma will continue to work, anyway, even if it's not the preferred form for arrays 19:49
timotimo m: OK 19:50
camelia rakudo-moar 47ddca: OUTPUT«5===SORRY!5=== Error while compiling /tmp/A05g_4Id2b␤Undeclared name:␤ OK used at line 1␤␤»
GLRelia rakudo-moar ba5494: OUTPUT«5===SORRY!5=== Error while compiling /tmp/ocaa_B24sY␤Undeclared name:␤ OK used at line 1␤␤»
19:50 telex left
colomon jnthn: gist.github.com/colomon/14fab50530ef7011b316 19:50
dalek kudo/glr: e26a731 | colomon++ | src/core/List.pm:
Fix for wonky combinations bug.

You've got to copy that array before returning it. Um, for some reason.
kudo/glr: b90cac5 | colomon++ | src/core/List.pm:
Check in golfed demo of the old combinations bug.

This is not intended to be a long term part of Rakudo. This is just for the purpose of debugging the insanity.
timotimo oops, the m: stayed
nine 68 failing now, official count! 19:51
colomon nine: was that counting my patch above? that fixed another file.
skids \o/
tadzik \o/
jdv79 nine++ 19:52
nine colomon: ah, I'm lagging :) Rerunning...
colomon nine++
laben what is this GLRR you all speak of?
timotimo moving the goalpost again :)
19:52 telex joined
dalek c: 33dc2ec | paultcochrane++ | lib/Type/Exception.pod:
Correct headings to allow die and warn to show up in search
 19:52
timotimo obviously it's not that "Great" if it only breaks so few tests
nine Great List Refactor Refinement :)
jdv79 what happens when the fail count hits 0j?
*0
colomon jdv79: ship it! 19:53
skids It prevented the slaughter of hundreds of poor tests :-)
TimToady --> horizontality
19:53 Peter_R joined
laben GLR Refinement? is that a soviet for Perl6 O.o? 19:54
colomon At some point someone was claiming that one of the modules completely messed up panda when you installed it. Anyone remember what module it was?
masak jdv79: yes, I was involved. I posted a gist about it. it's in today's backlog. 19:55
jdv79 oh...
timotimo laben: applied the patch and at the same time pushed some error reporting improvements i had uncommitted locally
masak jdv79: here it is: gist.github.com/masak/b694a4050f4e91bd254c 19:56
jdv79: note that this gist documents things the way I'd like them, not the way things are now
jdv79 oh, before bedtime reading - thanks
masak I think I saw in the backlog today that TimToady++ was doing something with .perl and cycles 19:57
skids TimToady++ tackling bears. 19:58
19:59 arafatkamaal joined
timotimo Your branch is behind 'origin/glr' by 165 commits, and can be fast-forwarded. 20:00
tadzik hah!
timotimo big ups
arafatkamaal gist.github.com/anonymous/3c294b28eda4e10aa7f8 20:01
nine 67 failing indeed!
20:01 darutoko left
arafatkamaal Boolean values like 'True' and 'False' can be variable names 20:01
timotimo arafatkamaal: yup, we don't reserve these words 20:03
tadzik so, on GLR my $c = self.bless(:$height, :$width, :$p, :$f); results with Method 'gimme' not found for invocant of class 'Seq'
does that ring a bell, or am I missing something obvious?
nine tadzik: there's no gimme anymore in rakudo's source code
moritz well, Mu has a gimme method that throws an error 20:04
20:04 arafatkamaal left
moritz so you should never get a "Method 'gimme' not found" error 20:04
timotimo for some reason my stage parse has gone above 60 seconds :|
tadzik so apparently I'm not actually on glr 20:05
moritz tadzik: but nom doesn't have a Seq type
tadzik: so something with your rakudo is *very* wonky
tadzik (╯°□°）╯︵ ┻━┻
20:06 mr-foobar joined
timotimo congratulations! 20:07
dalek rl6-roast-data: b23c606 | coke++ | / (2 files):
another glr run for 'today'
rl6-roast-data: fcf7109 | coke++ | / (2 files):
another glr run for 'today'
 20:08
timotimo ah!
i have a moarvm --optimize=0
nine skids: thanks for your work on push and unshift. It helped showing what we really need to do. 20:09
dalek ast/glr: 0fd48d2 | (Stefan Seifert)++ | S02-types/array.t:
Need to flatten explicitly in S02-types/array.t
 20:10
20:12 lucasb left 20:14 brrt joined
skids nine: np glad it worked out. 20:15
20:15 nebuchadnezzar left, nebuchadnezzar joined
dalek kudo/glr: 268ca08 | lizmat++ | / (2 files):
Make "done" a full supply citizen

Tomorrow, Jonathan is going to give yet another installment of his latest async presentation. We want people to use the new goodies as soon as possible and not have to change their async code again. So I thought it would be a good idea to rename supply-emit and supply-done to just "emit" and "done". This then interferes with the "done" from Test.pm (which I recently deprecated). It seemed like a good time, during this upheaval, to at least make this work in the GLR branch. So "done" is eradicated from Test.pm (please use "done-testing" instead if you must: you should probably use "plan" anyway).		 20:16
tadzik from done_testing to done to done-testing :) 20:17
another 200 PRs to all my modules :P
dalek ast/glr: d542bef | lizmat++ | S17-supply/syntax.t:
Make the changes of 268ca0802e6106de2880ddf in glr
20:19 virtualsue left 20:20 llfourn joined
jdv79 How about the idea of not polluting namespaces with exports? Test::done is pretty short and easy. 20:21
skids spares tadzik from PRs on Grammar::BNF.
nine jdv79: Test exports so many functions already. The one won't make a difference...
jdv79 or .done
20:22 xfix left
tadzik skids: <3 20:22
colomon lizmat: that was a harmless bug!
;)
jdv79 i guess in context. but in general seems kinda lame. 20:23
lizmat I *squashed* it!
tadzik skids: ooh
lizmat well, I guess I could leave sub done in there, just not export it
tadzik but now it doesn't work on nom anymore
lizmat but then people would need to change their test anyway
tadzik maybe we should just make them use plan()
skids Oh, nom doesn't have done-testing yet?
I'll do that then. 20:24
lizmat and changing it to done-testing sort of matches the P5 experience anyway (I'm told)
tadzik nope, still on old-done() :)
jdv79 really? we've already had the plan vs done debate in the p5 world.
tadzik we used to have done_testing in p6 world
then we changed that to done() because done_testing was RONG
20:24 llfourn left
lizmat well, that was before my time... 20:24
tadzik I don't remember the actual reasonong 20:25
or ning
there's always Test::done() :)
colomon Test:I-really-think-I'm-done-now 20:26
ugexe 1; 20:27
moritz Test::ichhabefertig
skids Test::mike-drop
tadzik or let's use ]
like bash
lizmat tadzik: done was implemented by moritz in 6a9a38f3b4d322a3ea93b
colomon skids: +1
lizmat on the last day of 2010
tadzik hm
I'm that old? :o
blogs.perl.org/users/byterock/2015/...first.html byterock++'s notes fromthe hackathon 20:30
lizmat yeah, byterock++
20:31 larion joined
dalek ecs: f3822a7 | lizmat++ | S99-glossary.pod:
Add GLRR lemma
 20:31
20:32 CQ joined
tadzik I guess that looking at roast commits would be a good way to write down GRL? 20:32
lizmat guess so 20:33
20:33 TEttinger joined 20:35 kaare__ left
ugexe this 2 hour time limit on macincloud is brutal 20:36
timotimo .o( is that a video game? )
ugexe its a OSX VM company 20:37
timotimo mhm
i wanted to build a website called "macontisch" a long time ago
which is a pun on "mac on table" in german
ugexe where im from it we had the phrase `crackentosh` which referred to the phenomenon of crack addicted people to always happen to have an apple laptop for sale (obviously stolen) 20:40
nine m: my @arr = 1, 2, 3; my ($x, @arr) = @arr; say @arr.perl;
camelia rakudo-moar 47ddca: OUTPUT«5===SORRY!5=== Error while compiling /tmp/lxhjImXpRp␤Redeclaration of symbol @arr␤at /tmp/lxhjImXpRp:1␤------> 3my @arr = 1, 2, 3; my ($x, @arr7⏏5) = @arr; say @arr.perl;␤ expecting any of:␤ shape declaration␤»
GLRelia rakudo-moar 268ca0: OUTPUT«5===SORRY!5=== Error while compiling /tmp/eDigtLlrG2␤Redeclaration of symbol @arr␤at /tmp/eDigtLlrG2:1␤------> 3my @arr = 1, 2, 3; my ($x, @arr7⏏5) = @arr; say @arr.perl;␤ expecting any of:␤ shape declaration␤»
nine m: my @arr = 1, 2, 3; my $x; ($x, @arr) = @arr; say @arr.perl;
camelia rakudo-moar 47ddca: OUTPUT«[2, 3]<>␤»
GLRelia rakudo-moar 268ca0: OUTPUT«[2, 3]␤»
20:41 rurban joined 20:42 colomon left, itz joined
dalek kudo/glr: c944e43 | (Stefan Seifert)++ | src/core/List.pm:
Fix list assignment evaluating the RHS too late

Fixes ($a, $b) = ($b, $a)		 20:42
laben tadzik: i have resolved the merge conflicts and pushed to my fork of panda, what do now? want a PR? 20:44
lizmat m: say Hash[Any].new( <a b c>, 42 ).perl 20:45
camelia rakudo-moar 47ddca: OUTPUT«Hash[Any].new("a" => "b", "c" => 42)␤»
GLRelia rakudo-moar 268ca0: OUTPUT«Hash[Any].new(:a("b"), :c(42))␤»
lizmat I think in the GLR, the latter is wrong 20:46
tadzik laben: yeah, sure :)
lizmat m: say Hash[Any].new( <a b c> => 42 ).perl
camelia rakudo-moar 47ddca: OUTPUT«Type check failed in binding key; expected 'Str' but got 'Parcel'␤ in block <unit> at /tmp/nCL8U3nmev:1␤␤»
GLRelia rakudo-moar 268ca0: OUTPUT«Type check failed in binding key; expected 'Str' but got 'List'␤ in block <unit> at /tmp/4oK1OX1jbq:1␤␤»
laben tadzik: github.com/tadzik/panda/pull/208
tadzik awesome, thanks! 20:47
laben i hope nothing broke panda/glr with all the commits
tadzik that's not all that much :)
I like it 20:48
I thought it'll be very mysterious and strange
laben yea it turned out i had to delete a ton of lines and that's basically it
tadzik awesome
well done :)
do you know if it's backwards compatible with nom?
laben our work up till now it sure was
our = mine + nine++ 20:49
lizmat m: BagHash.new(1, [ foo => 10, bar => 17, baz => 42, santa => 0 ]).perl.say
camelia rakudo-moar 47ddca: OUTPUT«([:foo(10), :bar(17), :baz(42), :santa(0)]=>1,1=>1).BagHash␤»
GLRelia rakudo-moar 268ca0: OUTPUT«(:foo(10)=>1,1=>1,:santa(0)=>1,:bar(17)=>1,:baz(42)=>1).BagHash␤»
skids lizmat: probably a flat or *@ where a **@ wants to be. 20:50
lizmat skids: spot on
tadzik laben: perfect :) I'll do some sanity checks and merge it tomorrow morning
laben i did not test the new version of panda/nom before merging nor did i test on the newest glr
tadzik alright
20:50 hudo joined
laben you made a ton of commits 20:51
O.o newest rakudo/glr got this: t/01-sanity/99-test-basic.t .......... All 71 subtests passed (less 4 skipped subtests: 67 okay)
smells like trouble
geekosaur skipped presumably means either missing prerequisite or known issue stubbed out 20:53
laben but it was working up till some hours ago 20:54
masak wohoo! it seems I accidentally made my first talk deliverable! :) 20:55
(in that it now contains 0 TODOs)
20:55 dalek left 20:56 dalek joined, ChanServ sets mode: +v dalek 20:57 colomon joined
skids 99-test-basic.t tests Test, intentionally todoing and skipping things. So there would have had to have been some finagling to not report skips. 20:57
20:58 Ven left, skids left
hudo it seems that $suchbegriff is not initialized 20:58
although it contains a string... 20:59
dalek ast/glr: 53c00f1 | (Stefan Seifert)++ | S02-types/nested_arrays.t:
We don't auto flatten nested Lists anymore
 21:02
21:02 bin_005 left 21:03 dwarring joined 21:04 bin_005 joined, colomon left 21:08 FROGGS left, brrt left 21:10 colomon joined 21:12 FROGGS joined
FROGGS m: my @a = 1 .. 5; @a[0,1] = @a[1,0]; say @a 21:12
camelia rakudo-moar 47ddca: OUTPUT«2 1 3 4 5␤»
GLRelia rakudo-moar c944e4: OUTPUT«21345␤»
laben tadzik: found out that on newest glr, panda need another slight fix, want me to add it to the merge PR or make another?
FROGGS nine++ # I was about fixing that earlier, but would not have found the patch me thinks... 21:13
tadzik laben: same one is fine
laben tadzik: and pushed, now to try bootstrap... 21:17
FROGGS m: sub bare_return { return }; my $s = "foo"; $s = bare_return(); say $s.defined 21:18
GLRelia rakudo-moar c944e4: OUTPUT«False␤»
camelia rakudo-moar 47ddca: OUTPUT«False␤»
FROGGS m: sub bare_return { return }; my $s = "foo"; $s = bare_return(); say $s.WHAT
camelia rakudo-moar 47ddca: OUTPUT«(Any)␤»
GLRelia rakudo-moar c944e4: OUTPUT«Use of uninitialized value of type Any in string context in block <unit> at /tmp/iKLXU0QTaa:1␤␤»
FROGGS hmmm 21:19
nine Official count: 64 spec test files failing
FROGGS when I make 'return;' returning an Empty, it will be a defined Slip when assigning it to a scalar
21:20 llfourn joined
FROGGS wow 21:20
64 is a nice number
timotimo it won't be as nice again until we hit 32 :(
FROGGS 42 is awesome too
timotimo mhh, true 21:21
FROGGS true indeed
timotimo mhh, defined 21:22
21:22 cognominal left 21:23 itz left, rindolf left, virtualsue joined 21:24 llfourn left 21:26 larion left
laben tadzik: why does bootstrap install the modules in <site> but panda lib and bin in $prefix? 21:27
lizmat hotel& 21:28
21:28 lizmat left 21:29 trnh joined
laben what's more, why does it not install the .moarvm files anymore for modules but does so for panda? 21:30
timotimo because precompilation is gone from panda 21:32
demon be gone!
foul, foul precompilation!
jdv79 interesting question - could the various vm's gen the precomp files themselves?
either if missing or on dep match fail? that seems like a better idea. 21:33
FROGGS jdv79: rakudo potentially should do that (automagically)
jdv79 *vms
i have no idea why i inserted a '
laben seems it's not gone away completely as panda does precomp itself, as i just said
jdv79 so just a NYI?
timotimo yes, NYI
jdv79 neat
timotimo precompiled files were supposed to only be a cache 21:34
laben FROGGS: and where would rakudo put the precomp'd file then?
FROGGS laben: into a CUR
timotimo but nowadays the presence of a "wrong" precompiled file causes execution to fail/bail out
that's obviously wrong
if that's encountered, the vm should just delete the precompiled file that caused the problem and compile anew
timotimo disappears into bed
laben CUR is the CompUnitRepo, right? where would it be located physically? 21:35
i mean computer-physically
timotimo it's the CompUnitRepo, but it's only a role
you're interested in CompUnitRepo::Installation i think
ugexe can you precompile just 1 file without having to precompile everything that is connected to it?
timotimo that's the one with the serialized database for installed things
laben this whole thing smells like incredible trouble for packaging
timotimo ugexe: depends on what you mean with "connected to it" 21:36
ugexe anything up stream might have been updated or deleted, and downstream would depend on it
FROGGS laben: the CUR decides how that happens, there is a database like CURLI and a P5/panda like CURLF
timotimo things a file depends on have to be compiled first, of course. and then they ought to be precompiled as well, i think
FROGGS m: say Empty =:= Empty 21:37
GLRelia rakudo-moar c944e4: OUTPUT«True␤»
camelia rakudo-moar 47ddca: OUTPUT«True␤»
ugexe right, but just having the vm automatically do it could then lead to unknown wait time
jdv79 no reason something like panda couldn't do it as an optimization, right ^H?
dalek p: 0a45369 | hoelzro++ | src/vm/jvm/runtime/org/perl6/nqp/runtime/Ops.java:
Make sure STAT_IS* returns false for non-existent files on JVM

Previously, non-existent files would return -1, which is still a truthy value.		 21:38
ugexe CURLI should be able to do it itself
laben what i'm thinking of right now is that: if you put the files/db in the user home, you risk wipes as (s)he could remove files; if you put it in a privileged position, you need privileges to rebuild them.
jdv79 yeah. might require a bit of sophistication ^H. 21:39
ugexe in the CURLI MANIFEST you just have to store the dependency build order (and provides build order), then it can rebuild itself
jdv79 where is this manifest you speak of? 21:40
ugexe also involves saving the hooks, %*RESOURCES, but i can see where this is going
%*CUSTOM_LIB<site>.child('MANIFEST')
*if* you have anythign installed to CURLI 21:41
jdv79 should i have one now?
i installed someting via panda
ugexe i dont know if panda installs with CURLI
jdv79 ah 21:42
i saw a statefile thingee
laben i guess you could install a cache in a standardized xdg directory, but do you imagine having to explain why everytime you use a module, it has to check the fs to see if it has been updated?
ugexe zef installs to a CURLI. panda to a CURLF.
FROGGS CURLI is not the default atm
aye
jdv79 ah 21:43
FROGGS CURLI though has currently bugs, and that's the reason it is not the default
jdv79 xdg is what now?
ugexe what bugs?
laben freedesktop.org standard
jdv79 what about windows? 21:44
FROGGS it is quite veeeeery slow when you install a lot of modules
pink_mist jdv79: standard for where config/cache/data files used by programs should be
ugexe I was working on that problem today. its because to-json is so slow. i was trying to sneak an outside to-json into CURLI with .wrap but it didnt work
laben well, since it already checks if it's on windows, having a dedicated WIN folder won't hurt 21:45
FROGGS and the solution to the slowness sits in a branch (jsoff), but this also is not bugfree (it crashes)
jdv79 i don't think the vm would have to check all the time. as long as the pkg mgr blows the precomp file away on a module update it should be fine.
japhb Wow, long backlog is long! Y'all have been rather active over there. :-)
ugexe this can output the projects.json in like 0.5 seconds vs default to-json of 2-3 minutes 21:46
japhb just got rickrolled by his music service
laben jdv79: the pkg manager would have to rebuild every reverse dep every time the module changes
ugexe tony-o json::faster
japhb What actually makes json::fast{,er} ... well, faster? 21:47
ugexe dunno, probably because it doesnt try to be clever
jdv79 laben: what? do you mean pre-comp? it could. but it could also just rm the ones directly related to the changing modules and let the vm fault it out
right now panda likes to fully reinstall revdeps so there is much room for improvement:) 21:48
21:50 rurban left
laben but where did the pkg manager put them in the first place? generally they use privileged access and unless you want to give it to rakudo too, it wont be able to rm/update them as it pleases 21:52
FROGGS ugexe: I'd be interested in your bare perl6 startup time when you installed hundred modules... 21:56
'perl6 -e 1' went to more than five seconds IIRC
ugexe oh for reading. i was talking about for *writing* the manifest after it gets huge
dalek rl6-roast-data: 2e63d46 | coke++ | / (2 files):
another glr run for 'today'
laben FROGGS: doesn't rakudo avoid searching for deps until asked (at compile time)? 21:57
FROGGS laben: might be true nowadays 21:58
tadzik laben: ...it shouldn't
FROGGS but then you have to wait the extra seconds once on first module use
laben why it shouldn't? if i say "perl6 -e 1", it shouldnt loadn any modules
FROGGS true, and it wont load any CUR code IIRC 21:59
tony-o japhb: it's json::faster is significantly faster, not sure what makes it faster than, say, json::fast or the built in to-json because i didn't bother looking at them
FROGGS so it wont deserialize MANIFEST.json
tony-o it's just a recursive method to convert hash/array to json 22:00
ugexe with 100 files (not distributions) in site/manifest the startup is still instant
laben the best way to make something faster is to make it work less
tony-o laben: github.com/tony-o/perl6-json-faste...s-jsonfast 22:01
laben i was talking about loading modules and what not, but congrats :) 22:02
tony-o ah 22:03
FROGGS tony-o: you do not plan to implement a fast json reader? 22:04
because that's what is the bottleneck for CURLF
writing json does not happen that often 22:05
tony-o the reader seemed like it was fast enough but i'll take a look at trying to make it faster
22:05 laben left
FROGGS use a large json blob and you'll see what I mean... 22:05
with that...
gnight everyone
tony-o later FROGGS
22:06 sno left
[Coke] Curse you Netflix. 22:07
japhb tony-o: Can you add the builtin to-json performance to that comparison? 22:11
[Coke]: What has Netflix done to you today? 22:12
tony-o japhb: i can add that, yes. it's roughly the same as the json::fast and json::tiny
both of those are comparable to the built-in
japhb ISTR the built-in was just directly taken from JSON::Tiny's code
tony-o i wouldn't doubt it 22:13
22:14 virtualsue left 22:15 khw joined
[Coke] japhb: SG:atlantis 22:18
22:19 bin_005 left 22:22 kinslayer joined 22:23 braintwist joined, kinslayer left
dalek kudo/glr: 817767b | (Stefan Seifert)++ | src/core/List.pm:
Fix list assignment involving sublists

Fixes ($x, @a[0, 1], $y) = 1, 2, 3, 4; We didn't handle the case that a finite List was contained on the left hand side. This caused the list in the middle to slurp all the remaining values, so the following scalar would end up empty.		 22:23
22:23 hudo left 22:30 RabidGravy left 22:31 bbkr joined 22:32 huf joined
dalek ast/glr: efb26f8 | (David Warring [email@hidden.address] | integration/99problems-51-to-60.t:
GLRify 99problems-51-to-60.t
 22:32
22:39 skids joined
japhb [Coke]: *chuckle* Watching it for the first time? 22:41
bbkr r: sub foo ($x) { say $x; samewith($x+1) if $x < 10 }; foo(1) 22:44
GLRelia rakudo-jvm 817767: OUTPUT«Can't open perl script "/home/steve/rakudo-inst/bin/eval-client.pl": No such file or directory␤»
..rakudo-moar 817767: OUTPUT«1␤Could not find dispatcher␤ in sub foo at /tmp/tmpfile:1␤ in block <unit> at /tmp/tmpfile:1␤␤»
camelia rakudo-{moar,jvm} 47ddca: OUTPUT«1␤Could not find dispatcher␤ in sub foo at /tmp/tmpfile:1␤ in block <unit> at /tmp/tmpfile:1␤␤»
bbkr what does this error mean?
dalek kudo/glr: 087a6c5 | (Stefan Seifert)++ | src/core/List.pm:
Fix Array handling in list assignment

While Arrays are Lists, too, they do swallow up all values stuffed into them, so we may not treat them like slices (which are Lists).
Fixes my (@a) = (1, 2, 3); broken by my implementation of sublist support. No other regressions appeared.		 22:46
nine Ok, that's it for today. Official count is now at 63 failing spec test files :)
Good night #perl6!
mattp_ jnthn: is a new 6guts blogpost coming soon? :) 22:47
[Coke] japhb: nah, but it was a favorite. pretty much just jumped to the top of my passive rewatch list. 22:52
japhb :-) 22:54
dalek rl6-roast-data: e3d5c20 | coke++ | / (2 files):
another glr run for 'today'
bbkr is it possible to get method handle within method itself? let's say I have recursive "sub foo { foo() }" and I don't want to repeat method name explicitly but to express "sub foo { call-the-method-i-am-in() }" 22:55
inner call may have different set of params 22:56
23:06 araujo left 23:07 araujo joined 23:08 araujo left, [Sno] joined 23:09 araujo joined
skids bbkr: $?ROUTINE maybe? 23:09
But if you want to do inheritance/MMD fallback look at callwith/nextwith.
23:10 araujo left
skids Erm &?ROUTINE 23:10
bbkr no multis, no inheritance
23:10 araujo joined
bbkr perfect, &?ROUTINE works. thanks! 23:11
23:11 trnh left
skids m: my @a = 1,2,3; my @b = <a b c>; @a = @a Z~ @b; @a.perl.say; 23:16
camelia rakudo-moar 47ddca: OUTPUT«["1a", "2b", "3c"]<>␤»
GLRelia rakudo-moar 087a6c: OUTPUT«["1a", "2b", "3c"]␤»
skids m: my @a = 1,2,3; my @b = <a b c>; @a = @a Z~ @b; @a[1,2,3].perl.say;
GLRelia rakudo-moar 087a6c: OUTPUT«("2b", "3c", Any)␤»
camelia rakudo-moar 47ddca: OUTPUT«("2b", "3c", Any)␤»
skids m: my @a = 1,2,3; my @b = <a b c>; @a[0,1,2] = @a Z~ @b; @a.perl.say;
camelia rakudo-moar 47ddca: OUTPUT«["1a", "2b", "3c"]<>␤»
GLRelia rakudo-moar 087a6c: OUTPUT«This Seq has already been iterated, and its values consumed␤ in block <unit> at /tmp/kLud5DYbFB:1␤␤»
23:20 llfourn joined
dalek rl6-roast-data: 43c8544 | coke++ | / (2 files):
another glr run for 'today'
 23:23
23:26 llfourn left
dalek c: 889e20a | (Tim Smith)++ | lib/Type/IO/Handle.pod:
Document .words and .ins
 23:26
c: 15603c8 | coke++ | lib/Type/Str.pod:
comb's :$match doesn't exist on Str pattern
 23:41
kudo/nom: 30d6fe7 | (Geoffrey Broadwell)++ | docs/ChangeLog:
Unsplit a ChangeLog entry
 23:45

Powered by Cro and the Raku® Programming Language.

Please report any issues / comments / feature requests as an issue on App::Raku::Log.

Thank you!